PROCESS PIPING SYSTEMS 381 be supplied to the pressure vessels on the process unit for washing out orhydrostatiz testing These comnections should be fiom the cooling-watersystem if the pressure in the system is adequate to supply vater to the top of th: tallest tewer on the unit; otherwise the connections should be to the fire water :ystem, Nonally vessols neod not be permanently counected to 3 source of water. If Permazent connection is made, it shoud be at the bottom ol the vessel and should be blanked off wien the vessel. is in operation. Air Piping. Most process plants have a plant ar system mot oily for use in the procesies but Lo operate tools, equipment, and instruments. ‘Where necesscry, the intakes of air compressor: should be designed to minimize the nobe level. Fiters should be proviced in the intake piping to reciprocating anc rotary air compressors when they take suction fiom the atmosphere. Filters wil sometines be necessary for centrifugal ir compresiors. Whena filter is not providee for a centrifugal air compressor taking suction fom the atnosphere, the intake Piping should be provided with a bird screen. Fiters preferibly should be-of the ; Feplaceable-cartridge type. Such fiers should have an open area not less then three times the area of the intake pipe. The oil-bath-type filter should rot be used with centrifugal air compressors Low points in the discharge line fiom an air compressor should be avoided because it is possible for lube oil to be trapped nd subsequently ignited. If low points are unavoidable, they should be provided vith drains When condensed moisture in air lines is undesirable from a process standpoint (which typically is the case-for instrunent air ccnnections) or the possibilty of moistue freezing exists, consideration should be given to providing an air drier Grum in the supply line near the process unit. The drum should be loccted where it will rot be exposed to heat from other equipment. Based on estimated future air requirements, the size of the drum should be such that (I) the velocity in the drum does not excted 15 fpm (0.08 m/s) during shutdown periods when maintenance equipment is being used and (2) the capacity be equal to at keast 6 percent of the free air requirements per minute during normal operation, In clmates where freezing is possible, the botom 18 in '450 mm) of-the dry drum should be insulated and heat traced. The d-um drain (or blow-cff) should also be traced or insulated. All blow-off connections should be installed pointing downwird so that any rust or scale blova out will aot endanger personel, Air piping should slope downward to dry drums or moistire traps, cr be hori- zontal, Branch cornections to air headers should te to the tcp of the ppe. Block valves siould be provided in all branch lines. Wher an air lin: is connected to process piping, two block valves, a check valve, and a bieeder shotld be provided. A second bleec valve should also be provided upstream of the check valve to test for tackflow th-ough the check. Consideration should be given to also providing a removable secion of line or hose in order to guard against inadvertent opcration Air for operating instruments is normally a sepirate system from the plant air system, and backup compression systems are often provided 1 increase in-service reliability. For process units, a steam-deiven compressor shculd be fumished to supply instrument air in case of failure of the main supply. Where plant air is the primary source, ant! the possibility of a power failure is remot, electrically driven compresiors may te used. In exensive insirument air systems, tre piping siould be a-ranged with header and subjeaders, such that groups of instruments mey be isolated from the systems without affecting tke air supply to all instruments. Block valves should be providede382 DIPING SYSTEMS. atthe instrument air headers in all branch lines to instramenis, Leads to individual instruments should be NPS ¥%4 (DN 15) minimum. As a rule of thumb, headers serving from 1 to 25 instruments thould be NPS 1 (DN 25) pipe size, and headers serving from 26 to 75 instruments should be NPS 2 (DN 50) pipe size. Steam and Condensate Piping. Process plants usually have twe ox more steam systems and an exhaust steam condensate system. One of the stean systems gener- aly operatesin the range of 100 tc 150 psig (690 to 1035 kPa) (low-pressure stzam), and another operates 2t superheated conditions and significantly tigher pressures (tigh-pressuce steam). The exhaust steam system normally operates at a pressure of less than 50 psig (345 kPz). The design problems astociated with these systems aie not all sinilar to those encountered in a central pover station, consequently a biief discussion on process plant steam piping requirenents follovs. The prindpal concern is to supply clean, dry steam to the equipment using it. In accomplishing this, itis desirable to connect all branch lines (except condensate ‘ellection.pcints) to the top of horizontal steam mains. However, if the line to a stam driveris at least one size smaller than the main apd the stesn has a consider able amount of superhzat, it may be permissible 1o make a cente‘line connection to the side of the steam main. With other steam contitiors it probably will be necessary to install a knockout pot or drum or 2 stean separatcr in addition to making the connection to the tor of the main. Pockets should be avoided in the line to the turbine. + Connections to exhiust headers should preferably 2¢ made to the top of the header so that the condensate in the header does not run back into the driver. In the steam line tc a steam criver, a block, valve(:) should be located at the diiver and be easily accessible for aperating purposes. A single gate valve is needed inthe exhaust line from each steam driver that does nct exhaust cirectly to atmo- sphere or ditectly into an individval condenser. However, valves reed not be.pro- viled where iwo or more drivers, which will never be shut down separately, exhaust to the same condenser. This exhasst gate valve should be installed at the driver so that the position ‘of the gate (ie., open or closed) will >e obviousto the operator whenever he is tequired to operace the inlet vaive. ‘Wherever steam is echausted to the atmosphere and could create such personnel hwzards as burns, freeziag of condensate on walkways, o- the blanketing of working area with a heavy fog, the line should be fitted with ar exhaust head and a drain toa sewer. Tae use of asilencer should be considered where noise misance is lkely. The fiexitility of steam piping should be attained through the tse of expansion bends and ebbow fittings. The use of expansion joints is discouraged except where the size and arrangement of exhaust lines prevent th: use of expansion. tends, which may be the case of certain steam exhaust connections to steam condensers. Particular attention shculd be given to the anchorage and support of the cornect- ‘ng piping. . ‘When recuired by the service, means should be available for purging process ecuipment with steam or inert gas. For example, each pressure vesse in hydrocurbon service should be provded with a steam-hose coanecton near the bottom £ not, permanently connected to the source of steam. However, where a permanent con- nection is made, it should be blinsed during operation of the unit ‘The stean supply for smothering, snuffing, service hoses, space heating, and auxiliary of 2rotective heating should be connected tc a source that will not be shut off during unit shatdowns or to a source thet wil not be shut off when the steam to a fiece of equipment sich as a turbine is stut off. For fire protection+ PROCESSPIPING SYSTEMS cata Purpcscs, smothering (or saultfing) steam usually is required for fied heaters and for relief valve cischarge lines Gondensate Renoval and Steam Treps. Concensate shuld preferably be di charged into an oil free drain system, but under no ‘circumstances should it be discherged into 2 sanitary sewer, Corsideration should be given to a condensate collce ion systemin installations whichinvolve a large numberof steam:raps. When Condensate is to 2¢ discharged to a cast iron or concrete sewer of a coreretn sewer box, tae hazard of vaporizing hydrocarbons which may exst in the sewer should pe cossidered. Also, to avoid damage to the coxcrete, the connection should be below the water level. If there is insuficient quantity of water for quenching, the condeasate should be first led to an atmospherepressure dain tank, Steam traps should be provided for the removal of condensate from collection Points in live anc exhaust steam systems, in partcular from condensate drip legs Grains on steam turbines, steam separztors, connectors, unit heaters, and termina fads of companion piping: Ail low points in steam lines, except steam companiot lines and the encs of long headers, shculd be prcvided with drip legs. It may alse be Recsssary to install drip legs at interniediate ponts on headers with long sestions at one elevation (ie., in addition to those low points at the end). When a valve’ installed in steam piping in such a manner that condensate can collect above the valve, a trapped drain should D2 provided above the valve seat Whenever posible, a steam trap siould be installed below end ciose to the cauipment pipelite being drained, but the trap should be easily accessible for Period inspectios. Each trap should serve only onc collection point. Where large Quantities of condensate are expected, cither condensate pots or condensate draive should be provided. : Drans from turbine shaft packing glinds and fom governor valve stem packing slands should preferably be connected to an open drain system. The drain lines and betdors should be of sulfcient size to preven: a backpressure builiup. Also, untrapped drains chould be provided at the lowes: point of the steam ead of each reciprocating pump and compressor. _—— Drains not discharging into a closed drainage system shouilddischarge downward and sheuld be arrenged so that rising stzam does aot create « hazard or condense Sn cauipment, such as a turbine or pump. The condensation of rising steam.on such equipment can create lube oil contamination. One thing that can te done to help eliminate this problem is to quench the.condensate. principal cawe of steam traps freeing is improperly designed discharge lines. Steam tap dischaige lines should be sicped for drainage where possible. In cases where. freezing is likely, no part of the trap discharge header should be at aa clevatioa above that of the trap dischaige. Focke's in the diicharge lines should be avoiced. Long trap-discharge lines, itt in heated enclostres, should be inou Hiet, miaP-dlscharge lines in heated enclcsures needto be insulited only ifyecessary {for burt protection, To decrease further the posibility of teezing, stam trap bodies shoulé not be insulated unless the followng circumstances make doing s0 advisable: * The trap is installed downstream of auiomatic steim control: that coule shut the steam off for long periods of time. * The trap is installed in a location where operators might be burned by the bare metal surfaces, * The trap is part of a heat recovery system where retention cf heat is important,cm PIPING SYSTEMS + The trap is installed tohandle exkaust steam condensate that comains quantities cf cylinder oi Inverted-backet and -hermodymmic steam traps, whch are conmonly used in prccess plants, are generally installed without strainers. Steam taps should be selected for 2 continucus discharze rate, which is the actual condensate rate muitiplied by a safety factor. A safely factor of at least 3 siould be used for inverted bucket type traps and thermodynanic traps. A larger safety facto: is needed for ‘xaos draining jacketed :quipment, and trap manufacturers shoule be consuited. In borderline cases offering 2 choixe between two traps sizes; the smaller trep is ustally preferred ‘Steam Compenion Pipisg for Auxillary Heating. The most commonly encoun- tered situatioas requiring auxiliary heating are as follows: 1. Piping in waich the fluid temperaxure could drop below the pour point or freezing point, and piping in which the fitid is subject to coagulation, excessive viscosity, or salting cut 2. Hydrocarbon vapor ad ges pipng where condensate formatior and icing will affect the safety and operation of the equipment, such as might be caused by che reduction in pressure that tekes place through’a control, throttle, or relief valve and 3, Liibe and «eal-oil eysioms for compressors and turbines Auxiliary heating is a6rmally not needed for freeze preventios and viscesity maintenance a equiprrent in intermittent service if tie equipment is draixed, Aushed, blows, or steamed out when there is no flowing steam, or ifthe equipment is for enough unidergromd to provent freezing. When required, auxiliary heating is wsually furnished by external steam companion piping (steam tracing). Other acceptable methods of heating piping and other equipment are internal stam trasing, steamjacketing, hot-water tracing and jacketing, md electric!racing. Details of tarious hea: tracing techniques are covered in detail in Chap. B6 ofthis handbook. Ic is desirable that each steam compenion line be coatinuous from the header to a trap at the end of the line without any vents, drains, branches, or deadend extensions at intermediate points. Each companion line should hav» a block valve at the upstream end and be arranged so that flow is generally downward, avoiding pockets as mich as posible arid kaving no section of the companion line at a greater elevation than te companion header. Live steam is preferred for steam companion piping in colier climates unless a lower temperature is required. In the design of the companion piping system, provisions shoud be made for the differentis] expansion between the traced line and the tracer. When the piece of equipment which is tc he kept hot is irregular in shape (such as ‘Taps, strainers, valves, and pamps), tubing must te used. The item should be sprally wrap2ed, stacting at the top and working toward the bottom. Several lines tobe traced nay be grouped irside a single covering of insulation if ther are to be maintained at the same temperature. CASE HISTORIES: CHALLENGES/SOLUTIONS © Process plantsofferthe pping desigaer some uniqite challengesnot feund elsewtere. ‘The combinations of denanding service requirements. end mechasical needs willPROCESS PIPING SYSTEMS 6.385 necessitate innlovaiive designs and soluions. Included herein are a fev practical approaches to proslem resolution: Challenge Installed Type 304, stainless steel piping and mating vessel nozzle NPS 8 (DN 200) Class 300 flanges were found to be inadequate for the specified 750°F (400°C) design tempersture and 440 psig (2760 kPa) design pressure hydrofining reactor service. ‘The reactor shell material and its other principal nozzles were constructed from low-alloy material weld overlayed with Type 304L stuinless sted. The reaccor flanges were allspecified with Class 300 flanges, vhichis acceptable forthe low-alloy flanges but not for the solid alloy Type 304L Manges. ASME B16.5 maximum working pressure for Type 304L SS Class 300 flanges is only 335 psig et 750°F (2510 kPa at 400°C) and clearly inadequate for the specified resctor design pressure. Solution ASME B16.5 downrates flanges in Type 304L miterial in elevated temperature service in comparison to the low-chrome ‘eactor Ranges and otter non-low tempera- ture grates of stainless steel. Attempts te rerate theinstalled Class 300 fiange based on the provisions of B31.3 for 10 percemt metal design temperature reductions for ‘uninsulzted flanges and the design procedures of ASME BPX&V Code Sec. VIII, Divisior 1 proved to be unsuccessful. It was ultimately decided to replace the existing Type 3041, flanges with Type 347 stainless steel Class 300 flanges, which permit ¢ maximum working pressure of 490 psig (6380 kPa), and as such, clearly acceptable for, the specified system design pressure Challenge ‘Two recent experisnees with installed spiral-wound gaskets with flexible graphite filler hare led to different but related ins ability proslems with these gaskets, which raised questions atout the standard ASME B16.20 covering these gaskets. The first experience involved Class 1500 and 250 spiral-wound (SW) gaskets fitted with inner retaining rings that suffered severe inward buckling at initial beltup. The second experience pertains to the gross iaward buckling of the inner spirals of Class 600 and ower rating gaskets supplied without inner “ings. Everts led to the ultimate removalof some 2(00 gaskets supplied bs 3 different manufactarers, which revealed that severe inward buckling bad occurred on about 12 percen: of installed gaskets actoss a wide rang: of sizes, Solutior Geskets supplied in both cases complied fully with ASME 516.20 requirements, which hghlighted inadequacies in this tandard fo: spiral wound gaskess. Efforts have betn initiated to work with manufacturers in their assessment of furdamental design considerations and with appropriate code committees to address apparent deficiencies discovered with this type of gasket. The buckling in both experiences386 >IPING SYSTEMS had occurred during the initial compression of the gasket, before :he flanges were pit into service While the root cause of the problem has yet to be identified, the buckling plenomenor is clearly related 1 the incompressible progertics of the flexible gaphite andthe tightness of the sriral windings. The experiences heve also revealed tnt the inner ring wieths specifizcd by ASME B16.20 for many gasket sizes are inadequate t> effective y resist buckling. Interim meastres have teen adopted by ASME B16.20, and it is recommended that users specify inner rings for all NPS 6 (DN 150) and larger SW gaskets with flexible graphite filer until more definitive measures ate identified to resolve the root cause of the bisckling phenonena Reference 1' provides more information on this industrywide coneern and overall cansiderations for gasket performance testing protocol: Challenge ‘Severe acoustically induced piping vibrations generated by high-caoacity presure- Jekdown valves have ledto fatigue failures at downstream piping braich conneaions wihin days o' their initial operation. One suck experienes involved « safety le control valve within an LNG treet gas unit to a flare reader sysem. The initial operation of this system ed to cracks at an NPS 10 (DN 250) branch connextion toan NPS 28 (DN 700) fiare header. The failure occurred after about 5 to 10 hours ofits inital startup and eventually kd to the branch connection completely brecking ‘axay from the run header. The Ietdowa valve was designed for amass flow rate ‘of about 383,000 Ib/r (-75,000 kar), with an upstream pressure of 620 psia (4278 KPa) letting cown to 30 psia (207 kPa) flare header back-pressure, ‘This and other experiences in the gas production, petrochemical, and other industries have demonsxrated that acoustic energy in high-capacity, gas pressure- reducing systems can case severe piping vibrations thatin extreme cases have led topiping fatigue failures within a few hours of commencing operation. ‘Sdlution Based on a statistical approach, criteria were generated to reflect maximum levels of acoustic power genezated by the pressure letdown valve before fatigue falure ‘was experienced (see Ref. 5). The ultituate solution involved the -eplacement of the installed pressure letdown valves with low-noise producing valves with labyrinth multistaged pressure reducing trimto avoid choked sonieflow condiions generated by the valve. Challenge Screre vibrations were found with the initial start-up of an NPS 2 (EN 50) nitrogen utiity line comected to an NPS 10 (DN 250) suction line cf a reciprocating compres- sor in make-gas service The nitrogen line was relatively flexible, and had ong lengths of supported piping. Detailed acoustical analyses were concucted with the man suction and discharge piping of this reciprocating compresscr, but this did not include the subject utility connection.PROCESS PING SYSTEMS C387 Solution Additional pipe sipports were installed to the nitrogen line, changing the natural frequeacy of the piping geometry and reducing ‘he response to the compressor pressure pulsation excitations. ‘Challenge An erosive slurry vas causing material loss at charges of direction in a convention- ally coxstructed piping system with elbows and tee:. Space coastraits did not allow for long-sweep turns. Solution ‘Dead-end tees were installed where the solids filled the impact area of tke tee, The abrasiv: solids then wore on themselves thereby p-otecting tte presstireboundary. Challenge Aheater outlet line was expected to ope-ate at about 1500°F (£16°C). The attendant expansion and stress analysis difficulties were magnified, since the process piping material was well into the creep range, Premature failure was expected Solution ‘The hot metal heater line was transitioned into an internally refractory-lired system near the heater outlet. The lower shell temperature eliminated the prosability of creep rapture failure, simplified the exsansion and stress analysis protlems, and reduoec system maintenance. ‘The previous discussion only touches on the multitude of challenges/solutions encountered in th: area of process piping. Reference literaure has documented some of the many valuable experiences encountered in this regard. REFERENCES 1. ASME Code forPressure Piring, B31.3, Process Pipitg, Americas Society of wfechanical Enghteors, New York, 1996 edition, including ASME B31.34-1996 addenda, 2. ASVE Standard B16.5, Pipe Flanges and Hanged Fiings, American Society of Mechanical Engineers, New York, 1996. 3. WEF. Bland, R.L Davidson, eds., Petrolum Frocessing Handbook, McGraw-Hill, New York, 1967. 4, J. Hirschhorn, Dynamics of Machinery, Barnes & Noble, Ine., New York, 1958. 5. V.A.Carucei andR.T. Mueller, “Acoustically Inducec Piping Vibmtionsin High Capacity Pressure Reducing Sysiems,” American Society of Mechanical Engineers, 82 WA/PVP- 8, New York, 1992.368 PIPING SYSTEMS 5. ASME/ANSI Standard B1.20.1, Fipe Threads, General Purpose (Inch) American Society of Mechanical Engincers, New York, 1992. 7. ASME Boiler and Pessure Vesel Code, Section VIM Division 1, Pressure Vessels, Americar Society of Mechanical Engincers, New York, 1996. 8. ASME Sandard B1C9, Factory-Made Wrought Steel Butt-Welding Fittings, Ancerican Socicty 0! Mechanica! Engineers, New York, 1993, 4. ASME Stindard B16.11, Forged Fuings, Socket-Welding aid Threaded American Society of Mechanical Engineers, New York, 1996 18. ASME Standard 81648, Steel Line Blanks, American So:iety of Mecianical Engneers, New Yort, 1997, 1. ASME Stndard B3610M, Welded and Seamless Wrought Steel Pipe, American Society of Mechanical Enginests, New Yark, 1996. 1:, ASME Stindard B1647, Large Diameter Steel Flanges, American Socety of Mectenical Engineers, New York. 1996 HL Bicklord, ed., Gaskets and Gaskered Joints, Marcel Dekker, Inc., New York, 1997. ASME Stundard 816.22, Nonmetdlic Flat Gaskets for Pipe Flanges, Axerican Socety of Mechanical Engineers New York, 1992. 1 LR. Payn:, RT. Mueller, and A. Bazergui, “A Gesket Qualifcatior Test Schene for Petrochenieal Plants: Parts 1 & 2” pp, 53-79, ASME PVP Vol. No. 158, June 1589. 16 LR. Payne, RT, Mueler, M. Derenne, “Specifying Flexibe Graphite 3ased Gaskets for High Tenperature Service (Qualfication Testing Protocols to Verily Performance),” resented at ASME PV&P Conference. Montreal, Quebec, Canada, july 1996. 17. RIT. Mueler, “Recent Buckling Experiences with Spiral Wound Fexitle GraphiteFilled Gaskets." pp. 23-34, ASME PVP Vol. No. 326, presented at ASME FV&P Confecencs, Montreal, Quebec, Canada, July 1996. 18 ASME Stindard B1620, Metallic Gaskets for Pipe Flanges: Ring Jon, Spiral Wound, tnd Jacked Ameian Society of Mechanical Eagineen, Nev Yor, 1986, incing ‘ASME B16.20b-1997 addenda. 19 ANSI Stardard B1.1, Unified Inch screw Threads (UN and UNR Thread Form), Amesican Society of Mechanical Engineers, New York, 1989. 20 ASME StmdardB16.4, Valves—Hanged. Threaded, and Welding Find, American Society (of Mecharical Engineers, New York, 19%, 24 APLStandird 600, Stee Gate Valves—Flanged and Buit-Wetting Ends, Bolted and Prssure Seal Bonns, American Petroleum Institute, Washington,DC, 1997. = 22 MSS Stancard SP-45, Bypass and Drain Connections, Manttacturer StandardizationSoci- ety, Vienm, VA, 1992 23 API Standard 607, Fire Test for Soft-Sected Quarter-Turn Valves, American Petrcleum Institute, Washington, DC, 1993. 24 APIStandird 394, Check Valves: Wafer, Wafer-Lug, and Double Flangec Types, American Petroleum Institute, Washington, DC, 1997 25, API Standard 598, Valve Inspection and Testing, American Petroleum lastitute, Washing ton, DC, 1996. 26. P.A. Schweitzer, Handbook of Corrosion Resistant Piping, Industrial Fress Tne., 1669. 21, ASSE Stardard A 1264, 1, Safety Bequirements for Workphce Floor ard Wall Openings, ‘Siatrs ana Xailing Systons, American Society of Safety Engineers, Des Plaines, IL. 1992, 28. D, Burgrem, Principles of Piping Analysis (Fst ed), C.P Press, Jamica, NY, 1977. 29. M.W. Kellogg Co., Design of Pipirg Sysiems, John Wiley & Sons Inc. New York, 1967. 30. ASME Boiler and Presure Vessel Code, Section IX, Qualtication Stardard for Walding ‘and Brazing, Americar Society of Mechanical Engincers, New York, '995.PROCESS PING SYSTEMS C389 31, ASME Standard B16.25, But Welding Exds, Americin Society o! Mechanical Engineers, New York, 1997 32, PFIStahdard BS-7, Minimum Length and Spacing of Welded Nozzles, Piping Fabrication Instkute, Springiale, PA, 1994. 33, PFIStandard E3-24, Pipe Bending Methods, Tolerasces, Procest and Material Require- ‘men's, Piping Febrication Kastitute, Springdale, PA, 1992. 34, PFIStandard ESS, Cleaning of Fabricated Piping, Pipng Fabrication Institute, Springdale, PA, 1993, 35, J.H.Bickford, As Introduction to the Design and Behavior of Boltet Joints, Maxcel Dekker, New York, 1997, 36, ASME Boiler ard Pressure Vessel Coda, Section V, Nondestructive Exeminaion, Amori- can Society of Mechanical Engineers, New York, 1995. 31, LH.Gary, and C.E, Hendwerk, Peiroleun Refining—Technologyand Econonics, Marcel Dekker, Inc., New York, 1975, 38 RE Johnson, “Specifying Plastic-Lined Piping.” Chemical Engineering, May 1982. 39, J.C. Wachel, anc C.L. Bates."‘Esceping Fiping Vibrations while Designing,” Bydrocarbon Processing, Vol.55, October 1976. 40, APIStandard 618, Reciprocating Comprassors for Peroleum, Chemical, and Gas Industry Sendce, American Petroleum Institute, Washington, DC, 1995, 41 APIStandard 6:7, Centrifugal Compresiors for Petwoleum, Chenical, and Cas Industry Senice, American Petroleum Institute, Washington, DC, 1995. 42, API Standard 610, Centrifugal Pumps jor Pevoleun, Heavy Duty Chemical, and Gas Indistry Service, American Petroleum fastitute, Wachington, DC, 1955, 43, APIStandard 614, Positive Displacemest Pumps—zeciprocating, Americas Petroleum Institute, Washiagion, DC, 1995. 44, APIStancard 6(1, Air Cooled Heat Excsangers for General Refirery Services, American Petraleum Institute, Washington, DC, 1997 45, APIStandard 60, Atmospheric Storage Tanks, American Petrobum Institute, Washing- ton, DC, 1994, vith 1996 addenda 46, APIRecommented Practice RP 520, Stzng, Selection, and Installuion of Pressure-Rellev- ing Devices, American Petroleum Institte, Washington, DC, 19. 47, APIRecommonied Practice RP S21, Guide for Pressure Relief are Depressurig Systems, ‘American Petrocum Institute, Weshingon, DC, 19°7.CHAPTER C8 CRYOGENIC PIPING SYSTEMS INTRODUCTION Nicholas P. Theophilos, Ph.D., PE. Manager Standards and Quality Ascurance Praxiir, Ine: Cryogenics (from the Greek “kryo-genikos,” meaniag cold generation) is the science and technology astociated with very lov temperatires. Depeading on one’s point of view, any temperature below ~20°F san be set fo establish such a demarcation. Here the ~20°F point has been selected because it normally represents the onset of embrittlement for ordinary carbon steels in typical structural applications, Cryegenics is not a separate branch of physics, snce it obeys all laws of ordinary physics. In fact, crfogenics is low-temperature physics. The reasons for its special treatment, therefore, are not because of its uniqueness as 2 science but rather because of the very special problems it creates: as a technology. These problems relate to embrittlenent of materials, lavge displacements (exoansion and contrac- tion), repid change of phase due to large heat fluxes (big delte T), and small latent heats of the fluids involved. In order to obtain a better appreciacion of the special cossideration: involved in cryogenic piping system applications, it was fek that it would be nezessary to review ‘he behavior of materials at cryogenic temperatures and the physical and thermodynamic properties of cryogenic luids. These considerations are covered in the sections “Progerties of Cryogenic Fluids” and “Materiab. Used in Cryogenic Piping Systems.” additionally, cryogenc piping system design is discussed in the sections “Piping Systems Design—Huids” and “Piping Systems Design— Mecharical.” From the strictly heuristic point of view of fundemental applications of scientific principles there ar> hardly any differences between cold box piping and all other types. Nevertheless, we are making a svecial topi¢ of cold box piping tecause of the conined spaces involved and the ccncepiual arrangements required to satisfy logisticelly workable and economically feasible process considerations. Such piping is discussed in the section “Cold Box Piping.” “The 2overage on cryogenic distribution systems as provided under the sections “Liguid Storage aud Conversion Systenss” and “Mobile Equizment System,” con- siders more than jest piping; it covers the functiona design philosophy ofcryogenic fluid stcrage and dstribution and provides quite an insight in-o the logistics of the entire operation, ‘With the advert of chip making, the need for ultra-high-urity inort gases has come isto clear focus, and industry hus responded to this need by developing suitablestorage and distribution systems These aspects of cryogenic piping systems are discussed in the last section. Naturally, the drive behind most tectnologies iseconomicin nature, ind in this 391ca92 HPING SYSTEMS respect cryogenics is no exception. This is certainly much ‘more so when it comes toliquid distribution because there is no other motive. Industrial gaies can certainly bedistributed in the compressed gaseous form. even in sulk quantties, if cosis are ct a consideration. Such economic aspects are discussed in the next section References for each section are at the end of each ection. ECONOMIC PARAMETERS OF CRYOGENIC FLUID DISTRIBUTION SYSTEMS Norman H. White, Engneering Associate Praxair Inc. Cryogenic processes are typically applied to commodity chemicds that exist as gaies at normal ambients, Such gaws are liquefied at rediced temperatures and are nomally maiatained at saturated conditions. The processing techriques generally involve both the liquid and gaseous phases and explo: the dramatic changes in physical and thermodynamic properties that occur with dranges of sate. Cryogenic ‘tenperatures are applied at cach step in the process of bringing these gases to the final consumer, including production, distribution, and storage. . ‘Typically, industrial gases ere found in mixtures in which some of the comporents: have commerzial value. Two of the most important examples of such mixtures are gates from certain hydrocarbon wels and the Earth's atmosphere. Tie feed streams are separated by liquefaction and subsequent fractiona’ distillatioa with the pro- duced stream: deliveredin either tie gaseous or liguid sate, Cryogenic distillation allows a wide choice In he degree of separation, ranging trom crude to extremely fine. It is a hghly efficient process with power consumed chiefly n refrigerstion lostto the envronment and pressure lost in the product streams, Ecoxomic consder- ations in the liquefaction and distillation process involve trade-offs between op- ‘erating efficiency and capital expenditure. No other method is as versatile or as effective as cryogenic separation of industrial gas mixtu-es for conmodity usage. The key toindustrial gas distribution and storage ope-ations is tke use of a cost- effective mettod that increases the product density. The -ransport or storage vessel is educed to manageatle dimensions in this way, The historical solution to this prcblem is elevated pressures at ambient temperatures. Indeed, high pressure cyl des and reo:ivers made to a variety of US. Department of Transportation (U3DOT) and ASME specifications are widely used today to store gases at pressures exceeding 2000 psig: (138 bar). These vessels are typicilly of singe-piece forged corstruction involving fabricetion and inspection procedures that enable then to safely operate at ultimate to desiga stress ratios of 3:1. In spite of these measures to achieve an efficient oackage, the ratio of vessel weight 10 preduct weigat is extremely hig. Therefore, the cost of storing and transrorting these gases in such vessels is very high relative to the value of the product ‘hey contain. Some of the characteristics of typical high-pressure receivers and cylinders are summarized in Table C8.1a and Table €8.Lb.(CRYOGENIC TIPING SYSTEMS e393 TABLE G8.1a -Typbal Speciations or Seumless Forged Pressure Vesses ay ASME Pressure Veseel Code. Sizes, capacitis, and design pressures 1 508 | 610 | 610 | 610 | 610 Min vall* in| 125 | 1.25 |1.303 | 1313 mm | 318 | 318 | 331 | 331 Max length r| so] 30] 23] a] 24] 24 mm | 9144 | 9144 | 7010 | 7010" | 7315 | 7315, Max water volume cut | 27] 27 | 342 | 342 | 602 | 602 liter'| 76 | 765°|969 "|, 969°] 41765 | 1705, Matera] 65:2) 70 [265] 70 | 65°] 70° asst ss : z. [F re Weight _ e _ - Unit weight thre | 232.4 | 222.4 | 2944 | 2944 | 2202 | 230.1 kgim | 331 | 331 |438.1-| 4381 | 343.9 | 343.9) Max vessel weight .« tb | 6672 | 6672 | 6670 | e710 | 5537 | 5537 kg | 3026 | 3026 | 3071.| 30:1 | 2516"| 2516 Design agit, ,ssc0-|, 2450, 332] 39 | 169 eis | aia" 184i. 249 | “2ea [327C394 PING SYSTEMS TABLE C8.1b_ Typical Specifications for Seamless Forged Pressure Vesstls USDOT speciication. Dimensions ana data for typical vessel sizes. —— vsbor i Speciation’ pei | 3442400 | 3AAx2400 | 344x-2900 | arax-aess | 2r-2400 | a7 2890 bar | 3AA465 | 3AACI65 | 3AAX.200 | 34AX-256 | TT-166 | 37197 Dimensions oD m) om 2 2 18 2 2 mm} 6) 39 559 457 39 | 58 Min wal = in | ost 086 ase 07s | ons | oun mm | 143 136 164 179 sos | ns Max engi? ein | 6-125 a 36 40 40 a m] 2a rig 097 wig | s219 | ais Max averoge Ib | Ue 5616 05s, sos | ass | sia veight kg | 5 2597 ma 262 | ro | ret Yom water me | 157 o 96 568 ns | oe volume titer | 44 2m D4 1609 | 262 | 200 2 Specifications areas defited in United States Code of Federal Regulations Tite 9 2 Length canbe varied to meet spec equiroments > Ipeiodes ent fing. Sours” CPladua Liquefaction is another approach that has been widely used for efficient stcrage and transportation of industrial gases. Gases typically errerge from ‘he liquefaction process saturated at approximatelr atmospheric pressure and from this poin: are transported and stored in eryogeni: vessels. These are typically double-walled ves- se with an ianer container design:d for the working pressure and temperatu:e of the product and an outer casing designed for ambient emperaturss and external pressure. In between the two vessels is a high-performance insulatien system which is wsually evacuated for the purpose of further enhancing thermal protection. Sruc- tual members to support the inner container and pipingto provide access to it are also located in the insulation space. Heatis continuously entering the vessel through the insulation, supports, and internal piping. This heat wil make the liquid contents bol, and the resultant zas must te removed from the tank if the pressure and tenperature of the contents ate tobe held steady. For tais reason, heat leak must be minimized The thernal efficicacy of the tank desiga is expressed in terns of the percentage of full expacity thot will be lost per day when the tank is hed at atmospheric fressure. This parameter is termed the nornsal evaporaion rate (NER) ‘and is produc: specific. Table C8.2 summarizes the specifications of typical tancage applied in vatious prodvction and distribution funetions Gaseous sate storage at high pressures and liquid state storage at cryogenic temperatures both have their placc in industry today. Therefore, itis important 10 Compare theit relative exonomics. The primary costs to be compared are those for storage vessel constructian, transpertation, power to acheve the storage stateand product loss. The constriction metiods used in forged high-pressure receiver: are very different from thove used in high thermal efficiency eryogeric temperature vesels. However, the resultant costs per unit weight may be considered equal for(RYOGENICPIPING SYSTEMS 0398 TABLEC8.Z Typical Specifications for Crrogenic Stonge Vessels 3 Model © DM- | DM | DM- | DM- | Dé- DM- | DM- 900°} 150% | 3000 | ooo | sao | 1100 | 13000 Capacty Warm Water git 960 | 1585 | 100 | 6020 | 9:80 | 1130¢ | 12300 liter 3634 | 6000 | 11734 | 22788 | 34°49 | 42771 | 50345 Net Liquid gel 900 | 1490 | 3000] 5880) 8900 | “11000 | 13000 titer or | sow | rise | 22058 | sse9 | snose | 42209 | Oxygen 1€00 cuft | 1040 | 171.0 |:3453 | 76 | r0a.4 | 12560 | 14960 | em oA 17790 | 2626:| 33277 | 39322 | Nivogn acoveue | s38'| i384] ams [ser | exxr | aoost [anion com 2203 | 36363] -t3ar_| 14301 | aui82 | 26916'| 31805 | 2 92s 71000 eu ft }-401.2 | 167.6-}°3375%p6615 | 108.3 | 1237.0.|-14620 | “um 2660-408 |~ 8871 ~) 17388 | 26e19-|-32515 | 38429 | nin 159 | 15-9 | 1540 | 262 | 297 | 315, | 360 a 48 | 48 |-433 |-20:| 99.|-96 | x0 | ftin 66° | 66°) 80 80 bs | 102° | 102 | em | 19°] a9-[-24-[-26 Weight ‘Tare 1000 Ib 95 | 105 | 155 | 27.9 f 10k | 43 | 48} 70 | 127 | we] 213 | m7 Orvgen: >, 100%.) 184.) 248] 44a} 999 | s24s | 1520,] 1700, vie wins 1 10 KB, | 82 | 112"); 200 | 38.0; | sea") | Nitrogen 1056 15 |) 156 [20:6') 357 -| 672.7] 910 *]> 1000 kg, Ta} °93'| 162° | 307° |) 439 “Argon 1000 Ib 20.0."}: 28.0: | $04 >| 96.3.) 1405 =}175.0.4.:201.0- so kg | 69.4 | t27-}-229 | 423 |. a7. |. 794.|.< 912 Design parameters . * ele NER (oxygen % per day}. 04 04 05, 03 0.5 0.25. 0.23, MAWE sia 250.,} 280} 20 } 50 :|--25) | 250 | “so bar 172} 172} 172 |. 172 172 172, 17.2" ‘Source Tayior-Wharton Div Harsco Corp.e396 PIPING SYSTEMS TABLE C8.3- Weight Ratio: Lading/Vessel Ctyogenic storage vessel Volume MAWP Lacing weight/Tare weight [a liters | psa | bar nN 02 AR 900 sor | 250 | x72 058 0.90 110 190 | 560 | 250 | 172 0.96 135 165 300 |. 11356 | 250 | 172 131 134 225 $880 | 22258 | 250 | 172 12 2.00 2aa 00 |. 33689 |. 250 | 372 182 2.29 280 noo | anes +]. 250 | x72 158 223 22 000 | 4920) | 250 | 372 us 248 3m Gas receiver J Volume MAWP ‘Lading weight/Tire weight cut | tiers | pois | bar Ne oz AR, 502 a70s~ | 2800 |: 193 04s. 018 0 461 |.1306 | 4000 | 276 033 047 oa “M2 969--} $800--] 379~ | - OM —|~ O45 - 019 20 fs 765-"~| 6667 |] 460} 040 O14 | 007 Source: Prasair, tre. oe ‘rough comparison. Therefore, storege vessél and transportation cost: for alternative systems can be compared by the ratio of thé vessel weight to product weight. Table C83 compares this ratiofor several products stored in typical gas and liquid stcrage vessels. Generally the wtal cost (apital plus operating) to liquey 2 product is ‘lover than that required to compress it-as a. gas to receiver pressure, Inceed, receivers are most often charged with gas pumped to presure in the iquid state and subsequently seated to ambient temperature rather ihan by ges state comprestion, ‘Product losset to be expected in verious liquid systems can be éstinated from the NER specifications given in Table C8.2. mo Considering these ‘circumstances, cryogenic liqaid phase storaze is thé’ nost ‘ecchomical approach in most cases. Gas storage should be considered where the -odpirement involves small quantities or long periods of noause or difficulty in disposing of the gas boiLof expected in a.cryogenic eys:om. The pipingused in cryogenic systzms obviously must meet the structural demands imposed by lew temperatures. Fron an economic point of view, the thermal eff. iexcy of the piping system mast be carefully considered ince the heat of addition, {o the system will ordimarily resul: in loss of product. There are two important, faciors of procuct loss involved in piping systeras that must be considered: refrigera- tiot required to bring the line to operating temperature (cool dovn) and steady state heat addition. Table C8.4 gives these parameters for uninsulaed lines, ines insulated with closed-ce] polyuretaane foam, and lines insulated with radiation shidds in high vacuum (vacuum insulation). .‘CRYOGENIC PIPING SYSTEMS 397 TABLE 68.4 Heat Addition’: Steady State and Cooldown Steady stat heat adition ‘Noninal ppe size NPS (DN) 3@ (100) 135) Bu Btu Beu Bu Insulation sett | warerm |e | waren | Jhe-A | vatt’m | forse | watt Uninsuated* 2630) aig |5330| siaz | 7870) 1562 | 10379 Polyurethane foam thickness Lin @5.4 mm) : 1s4.] 253] 243 | 329} 316 | 413 * 2 ia 6a nia) roa] I3a]-143 | 198]. 90 | 2a 3 in 062mm)" gs-/'22| u7 | 152] 146 | 192 ‘Yasin isuetion® oa |-0g| = 03} 20s x0 faa -amgh wind over Fosted insulated line." * Closed call poutine foam with PVC colt. ‘ Evorulated lamine radiation shields, SASIM BS Typel. + : * Schedule 5. , "ca alton duct inslton oly : fo” Source: Prexait re :398 HPING SYSTEMS PROPERTIES OF CRYOGENIC FLUIDS Theodore F. Fisher Process Engineer Praxair, Ine. Physical and thermodynamic proverties of cryogenic fuids consitute impcrtant dita that are needed for the desgn of cryogenic piping systemt. The following discussion, tables, figures, and references are furnished with this reed in mird. ‘Transpori property data are readily available for the more common pure cryo- genic fluids. The biblicgraphy at the end of this section lists sources of physical loperties and thermodynamic taoles and charts, which supply the detailed data Tequired forthe design of piping systems. Table C8.5M (Melzic) summarizes some ‘of the more important properties for a number of crycgenic fluids. ‘Thermodmamic. charts show pressure-temperaturephast-demsity (or specific vdume) and enthalpy {heat content, H) relationships ‘or a specie fluid under a .Natiety of formats. Entiopy data are often included but are rot gexerally required for the purposes under consideration in this chapter. Fig. C8.1 i 2 pressure-enthalpy shart for nitrogen, which will be wsed to illurtrate the behaviorof fluids in ransport systems, and various uses of pressure-temperature- velume-enthalpy (P-T-7-H) data. The most obvious use may be t» determine the density (the reciprocal of specific volume) of a fiuid which is being transported urder constant pressur> and temperature conditions. Referring the char, for caample, the density of nitrogen a: 101.5 psia (0.7 MPa} and 80°F G00"K) is fund tobe approximately 0.5 Ib/f? (8.C ka/m*). Other uses f the chat are illust-ated inthe examgles which ‘ollow, The dome-shaped curve ct the bottom left of the chart oncloses the two-phase (vapor-liqui¢) region. Nitrogen at pressure and temperature conditions to the left ofthe dome is a saturated or subccoled liquid. Conditioas to the right of the dome earrespond 10 saturated or superseated vapor. Vapor and liquid phases. coexist wihin the dome at a unique presture for a specific teriperature. This pressure is designated as the vapor pressure o° the fiuid at that tem>erature. The length cf the isobar (constant pressure) line between the sides of the dome is proportional to th: heat input which is required to :ompletely vaporize the fluid at he pressure and corresponding temperature. The fraction of the fluid which is vaper at a concition corresponding to a sperific point along the isobar is ecuivalent tc the fraction of its distance from the saturated liquid line, divided by the totel length of the itobar within the dome. ; ‘The pointat the verypeak of the dome represents a uaique condition of pressure and temperaure which is designated as the critical poiat of the fhuid. As nitrogen approaches 493 psia (340 MPa) and ~232.5°F (126.2%) from aay direction, all distinctions between the characteristics of vapor and liqud phases disappear. Lquid being warmed at a pressure above the critical pressure behaves zs a dense Tuid, which gradually approsches the characteristics of a hgh-pressure gas at hgher temperatures, without passing through any observable shase chanze. ‘The vertical lines or a pressur:-enthalpy char: correspond to a constant heat(CRYOGENIC PLFING SYSTEMS case content condition, Movement along these lines des gnates an Benthalpic or adiabatic (ao heat input or output) process, Flow through reasonably short or insulated runt of piping or fittings, or at a temperatureclose to thet of the environment, approache: isenthalpic behavior. When a ttuid sowing within a pipe is throttled through 3 valve, the change ir its pressure.tempsrature-phase and deasity relationships is »seentially iaenthalpic Tia fluid is initially a gas which is reduced in presture, its de:rease in pressure anc density may be accompanied by a chaage in temperature (Joule-Thomson effect) Tn most cases thé will be a decrease in temperature, as shown, for example, by nitrogen at 725 psia (5 MPa) and —153.7°F (170°K) following the 60.23 Btu/lb(m) (140kI'kg) isenthalp to about ~215°F (136°K) when throuted to atmospheric pres- sure. At conditions more remote from the eriticel point, th: temperatare change is less dramatic, A rise in temperature may occu: when the fluid is well above its critical temperature. Tais is shown by nitrogen ai 80°F (300°K) when throttled to 5800 psia (40 Ma) from a higher pressure, but it is more typically encountered with hydrogen or helium under normal processing conditiors. - ‘When a saturated liquid is throttled to lower pressure, an isenthalpic line is followed into the dome, indicating partial vaporzation of the downstream fluid, For exumple, if liguid nitrogen which i: initially seturated at290 psia (2 MPa) and ~251.€F (115.58°K) is throtiled to atmospheric pressure, the downstrean condition will be about 45 percent vapor. This vill be accompanied by a change in density from a2out 35.77 1b(m)/#t to 0.637 Ib(n)/fl (573 to 10.2 kg'nn"), or a laetor of 56 increase in volume. If nitrogen vapor is throttled from a point above the dome, it may pettially liquefy (retrograde condensation) or, within’ a very narrow range of conditions, pass firough the two-phase region before ending up as all vapor at a final low-pressure concition. ‘The large change in specific volune which may occur with throttling—and particularly vapotization—may necesstate a substantial incease in the diameter ‘of downstream piping in order to main‘ain reasomble velocties of the fuid. If the proces: leads to z two-phase downstream condition, there may be some slippage between the phases (icc., the relative Faction of the vapor end liquid iaventoried in the fiping may differ from that of thenet througkput). Two-phase flow character- istics can be extremely complex, and determinatien of this behavior is beyond the scope of this discussion. ‘An important consideration in the desiga of « cryogenic piping system arises from recognition that aconfined fluid cinnot inerease in speefic voltime, or change in density, when heated. Extreme overpressurization can reiult when a cryogenic liquid vhich is trapped between valvesis warmed, for example by heatleak. Con- fined nitrogen which is initially liquid at 101.5 psa (0.7 MPa) will [folowing the 43.70 Ib(mm)/f (700 kgim’) constant density line] =xceed 1160 psia (8 MPa) when warmed to ~2707°F (105°K) and approacit 14,500 psia (00 MPa) at ~100°F (200°K). This makes overpressure protection mardatory wherever entrapment of a cryogenic liquid or (initially) high-pressure cold gas is posible. Presure-enthalpy charts for other eryogonic fluids aro simlar to those for nitro- gen. Ctarts for oxygen and argon are provided as Figs. C8.2 and C83, respectively. Sources of charts for additional fluids are given in the bibliography at the end of this settion. The thermodynamic prorerty interlationshirs of these fluids are similar to those which have been illustrated for nitrogen. Rapid caposurs of a partially confiaed cryogeric liquid io heat may result in ‘overpressurizatior even when some outkt is provided. The possibility of c cryogenic liquid spill into.a warmer environment mast be considered in designing any enclosure around a piping system.° eur wo 0 feud eeu. «we mo eyo tad hoeod. aa oe ort ag tet et at ae uso sco geet, ewe BH et Gt poy rus vos eames ng cays ya's ic) cuos Sty ee sr ease sere Soest Ew woo aero sre “tro oo wo mo cue” 2is0% eo ato zo eco voz sco exo so ae oT fist 6 wore seoor = eber Sie ervapononso ont vor "fe atpt us 9S . wrt nye eee wet st wor sett eu > sang om yeas get seat eve siz re ett St te. eo Wor est syas sews ge ares woo wes ome setae ikon 10H) OND mew SHO wait sreuioW 18S wow saad, 2a (ousew) ig"80 STEW. c.400tio 99 capo 100 55200 wu 2000 BUY eeu ‘uy oer cen at aor srt ssc bathe ews ass ut Sone wee | seeme weme Tse ue to spo o0stop0 100 ao ‘sto ‘sto’ tooo too 100 oo 0 wo 150 seo reo on see, sei ow we otto eee wise” 900 eoe wee Se a (panna) epanya amtain ke: jaded aymmrtGpamran, pa peystina (ona) were sav e401(ea) sunsssud (2661 ‘8914 vumeayg “spi 24080) yo sonrode1g oeuEEipOWON,uouNDT pru0 OYE>UO.od ‘veeqoon, wid uogrucsd a peompondox) (sesX OO- WO = d) HoBOsIU 10) WERE Hrd. THD HUNT (EP) Ac THINS co ° (ea! aunssaud (Gur) AcTwHINA$0000 (ean) aunssaud wo o oz a oor ssa wrayer ‘spinys ausfokin yo sonsedosg sueeuspouony, ‘wounue pu opeouowog one Peonroniee) Coane WS UUO =) AEA 90) WP Ha C8 TAO (uM) Ad TWHINA e On ho Ce (ean) aunsssud (60) AdTWHLNA 6.403(ean; aunssaud (1661 AN sug temuarg spn swuatokia Yo sopsotos, ueuKpoMLOYY,uouUaT pu ojs2u0Ne, ‘wasqoonr wou) uexsnaad yn peompory) (Cai 09-900 = d) woRrE 29} WERE Hd E¥ TUN (aw) aunssaud > Gur) AaTwHNa » C404CRYOGENIC PIPING SYSTEMS 6.405 Certain fluidsrequire special consideration, Liquid carboa dioxide, for example, is notstable at atmospheric pressure. (This is indizated on Trble C8.5N. by a triple. point pressure above atmospheric). Loss of pressure on a liquid carson dioxid: system (or a high-pressure gas system: at sufficiently low temperature) will result in dry ice formation. Physical and thermodynamic property data on sryogenic mixtures are not readily available in easily applied chart or tabular format (as they are for pure ‘luids) Required design data are usually computer generated from complex correlatiors for specific cases. Knowledge of pur: fluid betavior will, however, provide a ‘awareness of many factors which require consideration in th: design ofa cryogenis piping system. -Anong the specific differences between the bzhavior of pure fluids and that cf ‘mixtures is a potentially much broader range of temperature end pressure condition, withir the two-phase region for the latter case and differences in éomposition between the individual phases and that of the overall misture. There is also 2 possitility of oneof the components freszingunder condition: wherein other compc- nents, may exist as gas and/or liquid mixtures. Although 1 component waich & below its norma. freezing point may have considerable solubility in tae liquid or as mixture, the possibility of solid formation should be atticipated when any cf the components s present in a mixture at a temperature below its triple point. This situation may result from the mixing of two streams, noithor of which contains 3 froze. component. ‘The designer should have knowledge of the corzbustion and physiological propei- ties o the fiuids being handled. Air and cther oxidants must be excluded fron piping transporting flammable fluids and from any enclosure (such as a cold box) inio which leakage may occur. Conditons which may concentrate flammable con- taminants preseat a danger in air separation. A number of cryogenic fluids ar: toxic, and all canpresent an asphyxiation hiazard, particularly within corfined areas. Venting of these fluids must take ino account the possibility that temperatur: differences from the ambient air may lead to unexpected localized ccncentratioa buildips which threaten the safety of personnel ind equipment. BIBLIOGRAPHY Sources of Physizal and Thermodynanic Properties of Cryogenic Fluids: AGA Gas Handbcok, AGA AB, Lindinge, Sweden, 1785. ASHRAE Thermotynamic Properties of Refrigerants, American Sodety of Heatng, Refriger- atingand Air Coiditioning Engineers, A\lanta, 1986.Tabular data and charts Beaton, CF. and G.F. Hewitt, Physical Property Data or the Design Engineer, Hemispher: Publishing, New York, 1989. CRC Handbook of Chenisiry and Physics, CRC Press. Boca Raton, FL. Jacobsen, RT., §.G. Penoncello, and E.W. Lemmon, Thermodynamic Properties of Crogent Fluid, Plenum Press, New York, 1997, E:tensive tabks and chars for the mgjor cryogens Langes Handbook of Chemistry, McGraw Hill, New York. Mathe:on Gas Dani Book, Matheson, Lynchurst, NT, 1980, Include: data for specialty gases, including toxicity, reactivity (Bammability,corrosiveness, ineompatbility, etc.) and safe har- dling procedures. Matheson Unabridged Gas Dasa Bcok; Matheson, East Ruherford, Nc: 1975. More extersive property data for these components.C408 MPING SYSTEMS Maxwell, .B., Data Book on Hydrocarbons, D. Van Nostranc, Prinveton,NJ, 1950. Tibular and chart daia for light iydrocarbors. National Stontard Refererce Deta Service of the USSR: A Seies of Property Tables, “temi- sphere Publihing, Washington, 1987. Comprehensive thermodynamic anc physical property ‘ables in volumes covering helium, nitogen, methane, cthane,oxygen, aiethylene, helocar- ton refrigerants, and rar: gases (aecn, argon, krypton and xenon) Pury, RH. aad D.W. Green, Chemnicul Engineers’ Handbook sixth ed., MeGraw-Hill New ‘York, 1984. “abular date and thermedynamic charts for many cryogens. Ruznjevic, Kusman, Handbook of Themodynamic Tables and Charts, Hemphere Publshing, Washington, 1976. : Reid, RC, JM, Prausnita, and -K. Sherwood, The Properties of Gases and Liguids, third cc, McGraw Hill, New York, 1977. Estimating methodsfor ptysical properties of pureffuids and mixtures. : Tubles of Themodynamic and Transport Properties of Air, Angon, Carbor Dioxide, Carbon Monoxide, Hydrogen, Nivogen, Oxygen, and Stean (orginally issied as NBS Circula1 564), Fergamon Press, New York, 1860. : TRC Thermoaamic Tablés—Hydrowarbons, 199. TRC Thernodynamie Tables—Now-Hy- ‘srocarbons, 1993. Thermodynamic Research Center, The Texas A&M University S)stem,, ~ College Station, TX. Mukiple volumes. MATERIALS USED IN CRYOGENIC “ hysical properties, compatibiliy with process fluids, fabriccbility, cost, and compliance with Tegulatory codes. ‘The subject of materials for cryogenic applications hasbeen generously treated in the technical iterature. For those irterested in additionalbackground information, a few selected works in tais area. appear in the references and bibliography a: the end of this section. The 2rime focus of this section, however, is cryogenic piping for the chemical process industry and commercial cryogenic distribution applications. Consequently, this overview will be deliberately limited in its seop: and coverage, and many materials’ (such as those used in aerospace applications) will net be covered. Thematerials trat Will be covered include ferrous alloys, nenferrous aloys, and nonmetallic materials. FERROUS MATERIALS Ferrous alloys most often encountered in cryogenic piping applicatons are ustally classified as ferritic or austenitic tyes. The terms austenitic ard fersitic refer to the predominant crystallographic phases fertite or austenitic, which are body centered cusic (BCC) and face centered cubic (FOC), respective y.(CRYOGENIC PIPING SYSTEMS C407 Ferritis Alloys “Most cf thie steels in common use are fertitic. This classification also covers steels which are martemitic 2s a result of heat treetmert. Low cos, ease of fabricability. and high strength via heat treatment are major ressons for their popularity. Ferric alloys, however, can exhibit a ductileo-brittle toughness tansition. The Charpy ‘Impact Test, by which measurements of energy apsorption, lateral expansion, anc ductile fracture appearance are made, is the mos common method of measuring this trensition, Hovever, the ductile-to-brittle trarsition temperature ir forritic steels can be influenced by a rumber of variables involved in the steel-naking process whict ‘contro, the levels of residual elements, inclusion shape, and heat treaiment tech- niques which in turn control grain size ind crystallographic morphology. compre- hensive review ofthese approaches is beyond the scope of this document. However alloying with nickel, reduction of sulphur levels, and reductions in gran size may be stated as the most popular approaches with ferritic alloys. Table C86 contains a listing of some of the most common ‘erritic alloy steels used in cryogenic piping Minimum service temperatures are inclided as we! as the applicable ASME specifi- cations. Tatle C87 and Table C8.7M (Metrc) containa listing of the same materials ax TABLE C86 ‘Typizal Ferrous Alloys Used in C:yogeric Piping Minimum ASUE® , Alloy temperature* | _speciftation Comrient CMa steal! 46°C (-S0°F) | SA-33°.Grade.1 | Aluminum tilled, fine srain practice DAM Ni steel 73°C (-100°F) | $A-337°Grade 7 | Aluminum tilled, fine “ - arain practize 3890 steel? ~tor'e (150°) | $4-383°Grade 3 |" Aluminum titted, nne arain practice 9% N steel! 196°C (~320°F) | $4333? Grade 8 -|. Aluminum killed fine grain practice 304 Sainless ste? | -254°C (-425°F) | SA-312 3O4L Stainless stee!’ | -254°C (—425°F) | SA312° 316 Sainless ste? | -196C (—120°F) | SA-31P 316L Stainless steeit | ~196'C (—320°F) | $4312 347 Sainless stee? | ~254°C (~25°F) | SA312" * Design mnimurr temperature for which maerialsnormaly suited witout impact vesting otter thar that required by mateial specification, " Femitie steels. rueniti steel _Conesponding ASTM speciation ste ABD and A312chew pus eeey aa sonoeid ured our ‘unum ‘p32HeaHION “EE-Vs wonkoypods ap ySy UH Paves Se copes UBUD ° zu > oe oe oe - wea — SB “8 09 os 99 06 Lt ae WS SSO1UIIS [pe grese aes | oie se oung se- | eat as OTE 88 ouFeg OTe se owes -. oy ge SB Le ae WS ssoqunens “I9TE ° ¥ = 9s 9 161 see oie al. Eo 6 06 = sr 8 a 1a ae vs | sawes oie wor sv owes | poesvouns | oo F a ca - | toc a poe POE $2 auteg ‘POE Se eurES 09 ov a os Lt ae VS ‘SSOTUIEIS THOE ° zt st € ose sor We aL, 8 6 06 su a 8 LY zie vs | somes yo vt ar sc a ser ont oe- | gape est cs le a 6 su 1H cic vs | gow i as = sr wz a oz ost) eapei0 we as 96 wt sl ool La ‘€f€ VS 2191S IN HAE = = oc o” st oor | Lape 7_ _ 8S ost se so Le ‘€c€ VS 9998 IN MT = os- | Tapeip of al? 0c ss a €ce WS Pas UW (.-1)nLa) (® = a) pads sony Aunmpoos | uze | wsuans pera | WA vampersduas, | aWsy recon), wou, — | roadur | vowe8uora | resuo seo" | 2usun ddrouy, rena) stony enessoa je sopedoig roondya PUT Teonseyseyy INIKL £ By a RIVE cave‘uj munesg. anos ‘sinyerdinsy woe & LAL, choy we pesy oe eae BY pmo pw pate sojeand vies? uy pom wmtzuange‘paneaoN + REEYS woHEUDAds ANSY- MI PAS aNA MCHA “0 eat or tee Fave Teo + OST oon a Sarevs | = seems 29 gtese sues | ore se cures i %61- ‘WIE dL por gTese suws | gre sv cues a we zie ws | ssouers i9ic 0 3 wr- | 9c pas st 2% a ZEVs | © ssomms orc ‘or, se owes 6 ¥e7— Atos an po 08 Se ouses Se, Ly ZTE WS | somes “THOS 1 £8 18e— 40 dL. pas wor lt Di zie vs | -ssomureis por 98 166 © 96t- 89p210) | vor 19 = oe seevs | aoasin x6 98 985.5 Tor £2781 ror ‘us Lt EE VS ene = ote ee co - Lt eee VS WAT ys op L9p2D, ct Le seevs | ras uN OD “pads fouy woreda pawaducs | ainsy eunoqy, story sania jo topiary hn DH PHOS HNIEL, (ONAIN) ILOO ITOVL 409e410, BING SYSTEMS Teble C8.6 with typical data, including some data not found in the ASME spedfica- tens. Included in Tables C8.7 and Table C8.7M are mechanical properties such as stength, impact, and elongation, «s well as thermodynamic data such as thermal expansion ard thermal conductivity at cryogenic temperatures (boiling point at 1 atmosphere). Austenitic Alloy Steels In addition to providing data on the ferritic alloy steels, Tables C26, Table (8.7, and Table C87M also contain a lising of austenitic stainless sieels which are Lkely tobe encoun‘ered in cryogenic piping. Most of the auscenitic alloy steels used in cryogenic pipng are chromium-niccel stainless steels of :he AISI 3)0 type, such as 301, 304L, 316, and 316L. Other stainless steels classifizd as martensitic, duplex, ani precipitation hardening also zxist; however, the preceding dloys are nost conmonly used in cryogenic piping for chemical proces: and distrbution applica- tioas, - Acmrajor consideration in the use of the 300 Series staialess steelsis the improve. ‘ment in toughness properties they provide and elimination of the sharp ductile-to- britle transition found n ferritic steels. These alloys have been ised to cortain' and distribute iquid hydrogen and helium. Consequently, while the austenitic siain- less steels were originally developed for corrosion resistance, their toughness and excellent fracture properties are what led to their selecticn for cryogenic piping ap- ications. ; : nO the 300 Series alloys, the AISI.304 compesition is the most popular as. ‘measured by ionnage. It should be noted that there is a oreference for AISI 316L. in the electronics industry for the distribution of high-purity gases that are free of; articulates, Furthermore, piping and tubing used in this application are frequently: eleetro-polished, and the trend is for an increase in electro-polishing for ultra-high unity applications. Cast versicns of the somnion austenitic stainless steels may be used if valve applications exist. CF3, CF3M, CF8, and CF8M are cast equivelentsof 304L, 316L, 304, and 316, -espectively. Note thet the cast stainless stzels have teen optimized forcastability, and incressed delta ferrite levels may adversely affect toughnes at cryogenic temperatures. Similar efiects may be observed in austeritic weld fller meials, and both castingsand weld filers should be well characterizei and qualified rior to, use in cryogenic piping sysems> NCNFERROUS ALLOYS Nonferrous alloys encountered in cryogenic piping are usually of the aluminam, ‘cup‘ous, or nicke] families. None of the three alloy families exhibit ductile-to-britle toughness transition due to the FCCcrystal lattice. Common nonferrous alloys wed in cyogenic piping appear in Table C8.8, Table C89, and Table Cé.9M (Mettic). Aluninum Aloys Common aluminum alloys used in eryogenic process pipng do not represent the gamut of whatis availabe in aluminum alloys. The high-strength aluminum alloysCRYOGENIC PIPING SYSTEMS, TABLE C8.8_ Typical Nonferrous Alloys “Jsed in Cryogenie Piping : Minimam * ~~] > ASME! Alloy Tempers temperaure spec. 1100 Aluminum 0, EM 254°C (-452°F) SB 210 3003 Aluminum, 0, uz ~254°C (-452°F) $B 210 5052 Aluminum, 0, ER 254°C (-452°F) $B.210 5083 Aluminum om | -2ec earn |. Se20 $086 Aluminum, ori: | -2ec(asrR |” sa 210 6061 Aluminum 256 (as2"F) = |. sa 210 Copper (C10200, €12200) 5 ‘Copper-nickel (70600) _ € $8467 ‘Copper-nickel (C71500) EE Boece S|. spse7 ‘Monel 400, Al-Ca alloy 198°C (-325°F) _ |. SB 165 * Design minimun temperature for which iatéria is normaly Stable without inipa testing otha = fan the required by material specication | Coresponting (STM seseations ae 37, B16, BzI0and BAT. commonly used in aerospace aplicaions are rot used in the chemical proces} = industy Beaute ofthe lack of ASME Code coveigeoroter consideration such = as-corosion. Pere Aluminum compositions of the 5183 and 661 types constitute the highest. - strenghh alloys wed in cryogenic applications, and‘tempers of these alloys ‘are. = Suitable at temperatures as low as —452F (—254°C). From the toughnessstandpoint, 5083 would be preferred. However, prolonged exposure to temperattres as high as 152F (66°C): faring ‘thaws can result in both areduction .n corrosion resistance and toughness. For low-strength applizations, the 3000 series are also ased. 5 Copper and Cuprous Alloys et Most of the early. work in the cryogenic industiy was accomplished through the “use of copper process vessels, piping, and tubing, Aluminum has extensively replaced * copper and cuprous alloys in the fabrication or air Separation plants However, copper and cuprcus piping are still extensively used in piping and tubing runs from cryogenic tankage for several readons which includ» eade of fabrication (ag., brazing, solderng), resistance to ignition, and combusticn resistance in oxygen-enriched systens. It may be noted that orbital are-welded stuinless piping systems are increas- ingly replacing cxprous piping for the delivery of particulate and contaminant-free electronic grade gases.‘OTe ‘SLs 918 suontoynads WISV “Areansadsox supnipsoo pouapiey pue os aImdysep HI pur § 8I97>| UL « s se os 9 0s sit Ste st se 3s o sc 08 TE | soras Oot [340K a 0 te se ste- ste yao u 6 ol oF o + Ta] toras INE C€/0L | Ste soe vad 9 | 6 aw or oF sa | ras ENED 01/06 st os Se 9” vt we ep (oozid oer ce 95 s or te ta | sews | vv) ssa z ae] 9s oe 2p OL 66 or or a ov sp La | orz-as, | wonurmmyy 1995 we see | Het soz | 196 se | csr- ze ‘0 e ver Her see | Hutt sit | wep see | ia | orcas | wnuwny 9905 + s See sez seo | csr 0 89 er vor o7 | ucer siz | um see | au | orcs coos ‘ HE sey.| Hote suv | 9s soz | zop— zH‘O. se ver HB -Sct,| Hose sie | Hop ser | au | oras | umawnry zoos Ey ‘sev 30 see | sr aH’ 6 HL Sse | Heat. 9° | Hz sor | ay | orcas | umuumy soo ot $98 ty zm '0 ser HOL Sob sp | reas | uimanunry cont (a.-¥-WaLe) ") (%) (a) “eds Aory Auanonpuos ‘vowuedya wpduans ea ‘dwet | .aNsv reunouy, rewroq, yeeduy | uoneduorg faro49, _| sSfony sromayuoy, jo sorsodoug ropes PUR ToRUONDOYY TOOKKL GED ATAVL caiz‘suy yexeg snes, in OL PO T2919 set om montenosie wLsy Suptodes, sO : © 8 soras or 1aU0Hy = sor- [Pi sib yuu 6 or ia | wras IN-9O 06/04 aot | 90, vad sy a sbi | isi-as I-79 0106 ocr m6 OE 861- (ooze cae ee waa | sias | ‘eoz01o) seddo3 ar: ec a ut a 2a | ore-as | wnummry 1509 GLI Sect | vst THO ger & HLIL Set Ta | tas | amumnry ops “ 6 S65 ¥st— , zu‘ at ez Hest sere ebr: | oreas | aaumnry esos zt HOLE s¥ce vst- zee‘. oer « Hose sssv. Bi | orcas. | wnupunry eops an 6 s9! ¥s0= a0 ost w H9gt Sir’ za | orgs: | wnumnry cope os? 6 sss ¥50~ Caray oz @ Hast SHE ax | orcas | anummry oort ewan) “| G00 x 87) | ‘sads fou Axapenpuee ‘ropa aiady, | ansv PULL, TUL siete askory snonagnegs Jo eanearions resektar Pwo TeoqueRaayy Te=AL, (OMMOW WNO'ED TIVE carscat4 TIPING SYSTEMS ‘Nickel Alloys ‘While nickel alloys could be used in eryogenic piping srstems on the basis of their mechanical sroperties, their high cost generally precludes their use. One najor exception is he use of Monel,® a rickel-copper alloy, which may be used in oxygen systems to rrinimize ignition tendencies where there is soncern ofimpingement or the potential for high velocities. NONMETALLIC MATERIALS Although the bulk of the materialsused in cryogenic pipelines are metals, nonnetal- lic materials nave criticel functionsin cryogenic pipelinecomponents such as valves and insulation, Numerous nonmeallic components have been used in cryogenic Pipeline appications, and it wouldbe beyond the scope of this section to cite them v al. A brief listing of some of th: more common nonmetallic materials used in pipeline components is found in Table C810, Table C8.11, and Table C811M (Metric). - TABLE C810 ‘Typica, Examples of Nonmetallic Materia Used-in Cryogenic Piping 8 Application Material Gaskets Durable, (bonded or compresed asbestos | Grafoil (exible graphite) = Insulation (Biber) Mineral woo! Fiberglass =; | Insulation (foam) Polyurethane : Styrofoam * Foamgless(R) 2. | Insulatios (powder) Perlite Vermiculite Insulation (sheet) < Mylar ‘Aluniauni opacificd paper Insuiatior (support brock) Transite Micarta ‘Valve packing, seals, and lubricants | Viton KELF Glass-teinforced Teflon Fluorclube_ ‘Source: Praxalr ia.CRYOGENIC PINGSYSTEMS cats TABLE C811. Typical Mechanical and Pliysical Properties of Nonimetallic Materials “enaié [Moats] emt | Tews | Sea sang |"taty | Speete| expan || contoy ae GS" | “Gad” | ary | Gar 10-9] ret") ee Bee) [rat Gor paar anal) i ‘Gnd Gee ie) asa | 2m | 1a | am om [a Merl vol ernst | 02] asa} 0 ‘igs fi ition 5m | “ome 05.08 Verisalite powde asain wef | ode ‘Myr stl» as sf 19 |: ‘Through aictness conduetwity, Souree:” Prexar Ine. plssis ob TABLE C8:11M Metric): ‘Typical Mectanical and Physical Properties of Nonnete Meena) Peat Mesinie . law cen Hepsie | Nous], Tuewa | aeaw Sue EEA | || ae |e [ay Dunble (asbestor gaket maserial) |.27~75 |, . - Gatton emmy fer | ae | Ser wari | > waa] Ts ; “Fibemlass fiber insuktion 344-3400 | 57,880. ay Pet powder insaion| Verniuiteromde i a la been” um | wa | ap am 28 | hone. Aluninum ose oper abet | No loner Ilon valabe “Trant innntion bk Ne loter ea * Through thickres conductivity, Source: PraxairIne,cate PIPING SYSTEMS Again, compatibility and mechanical and physical properties must be considered in the selecion of noametallic materials. With respect to comoatibility, iquid onygen is the commercial cryogen of greatest concern because of its large tsage. ‘Typical tests or experimental parameters covered in tke selectior of nonmetallic materials for oxygen service include aufoignition temperature, hea: of combustion, inpact test, and oxygen index. Materials that are compaible with cxygen generally heve high auoignition temperatures, low heats of combistion, hig energy absorp- tion in impact tests, and high oxygen index values. Reference’ gives advice roduction Consequently, many common materials turrently used in cryo- genic 2iping systams are likely to be fhased out over time. Rerent legislation mandates that material safety data sheets (MSDS)be obtained for all industrial substances that are ia industrial use. This obviously includes the metalic and nonmetallic materials used in eryogeric piping. Failure to comply with the requirementsand implementation provisions of the original legislation can have serious consequences for fabricators. Equipment used in oxygen services 1as additional requirements such as cleaning and velocity limitations which must be considered. Oxygen equipment must be cleaned to elimizate contaminants such as hycltocarbons and metal particulates, which could serve as ignition sources. Reference 3 provides a general discussion cf this issue. In addition, ‘velocity limitations may be placed on certain classes of inaterial to preclude ignition by particlz impingement. See Reference 4for velocit limitations and d:sign considerations peitinent to oxygen sjstéms. ‘Other industral gases that-might te encountered in cryogenic pipeliries have specifi hazards «sociated ‘with them (ie., fammability, toxicity, ete.) These are summarized in MSDS forms available from induitrial gas suppliers. REFERENCES. 1, “Guide for Evahating Nonmetallic Materials for O:ygen Servite,” ASTM Specifcatior 663, Amecican Society for Testing and Vaterals, Pulladetphia, PA. 2, “Design Handbcok for Liquid Fluorine round Handling Equipnent;" Technical Report No. AFRPL-TR65~133, Air Foros’ Rocket Propulson’ Laboratory, Edwards: Air Force Base, 1965. : 3.16. Hist and Git, Ady #A\ Survey of Compatibitty of Matedais, wit High Pressure ‘Onysen Service,” Cryogenics, vol. 13, 1973, pp. 325-335. 4. “Cleaning Equipnent for Oxygen Service.” CGA Panphlet G.4. 1,Compresse¢ Gas Associ- atior, Arlington, VA. BIBLIOGRAPHY Handbcok on Mauerials jor Superconductixg Machinery, Technica Report MCIC-HB-04, Metal & Ceramics Information Center, Batelle, Colunbus, OH, November 1974. ‘Metals ‘landtook (desk edition), American Society for Metals, Ceveland, OH, 1985, pp. 170.2420 34 Reed, R.P., and Clark, A.F. Metals Park, OF, 1983. ‘Schwaruberg, F. R. Cryogenic Materials Dita Hendbook, vols. 1 and ll, Techical Report AFML-TR-G4-280, Wright Patterson Air Force Base, D4, July 1570. “Materials at Low Temperatures.” American Society for Metals,cae. TPING SYSTEMS PIPING SYSTEM DESIGN— FLUIDS M. J. Lockett Corporate Fellow Praxair, Inc, PIPE SIZING CRITERIA For all cryogenic fluids, except onygen gas, pipe sizing is based oa pressure drop censiderations (for oxygen gas, seethe section “Oxygen Sas Piping”), The pressure dop criteria presented in Chap. B8 ate applicable to cryogenic fiuids wher the fitid is either in the liquid phase or gas phase flow regime. Heat transfer into a saturated cryogenic liquid ot a érop in pressure can cause a portion of the liquid to flash inte a gas (sce the section “Properties of Cryogenic Fluids”). A two-phase (gas-liquid) flow condition must then be considered vhen determining the required pipe size and pressure head requirements (sce the next sestion). Two-phase flow will resut in larger pressure crop losses or a larger pipe size requirement than will liguid phase flow. Therefore, whenever possible, the fluid should be maintained in a single-phase flow condition. Th: liquid om be suacooled to prevent fiashoff due to pressure decreases such as zcross thrattling valves. Use of pipe insulation car reduce the heat leat into the piping, thereby ako reducing flashoft. TWO-PHASE FLOW Fer prediction of the pressure drop experienced by a two-phase cryogenic fluid fiowing along a pipe, the usual rractice is to divide the pipe irto sections for calculation purposes (see Fig. C84), At the beginning of @ given pipe section, . ig ae i » prado h h+Ah FIGURE O84 fipe section.(CRYOGENICPIPING SYSTEMS cai9 conditions are assumed to be known. The pressure drop over the section is'then calculared based on the conditions at the begining of the section, and so the pressure at the end of the section can be determined. The heat leak into the section is estimated from heat transfer correlaions. An exthalpy balance over ‘he section then alows the exthalpy (h) at the end of the sortion to be calculated (potential and Kirti energy changes ver the sertion can ually be nigleted), Knowledge of the pressure and entialpy at the enc of the secon allowsnew conditions to calcula‘ed from tke thermodynamic properties of the cryogenic fluid. Such condi- tions include the temperature, vapor fraction, and physical preperties of each phase. ‘The caculations are then repeated for the next ection, and soon to the end of the pipe. ‘Two-Prase Pressure Drop The toral pressure drop for two-phase flow consists of three parts; the frictional, gravitational, and accelerational pressure drops, s9 that, ap apr (#2) (2) 2 Pe) 4 (Be (C8.1) (2), +2) en where p is the pressure, pdl ft (Nm™)and z is the distince along the pipe, f(a). Mary correlations have been propcsed to prelict the fri:tional pressure drop (épp/d:) and they are discussed in Chap. B8. The 2arliest comelation, which is still widely used, is that of Lockhart and Martinelli? To use this comelation, the steps are: 1. Calaulate the Reynolds nimber for each phasé flowing abne in the pipe. M.D Re, = —— (C8.2) a (€8.2) Reg = Me2 (C83) Autc : where M;, Mg = liquid and vapor flow rates, Ibs"! (kg s™ AD cross sectional area, ft (m’), and diameter, ft (m) Hy ie = liquid and vapor viscosities, Ib °! s-! (Nsn-?) 2, Calailate the frictional pressure d:op for each phase lowing alone in the pipe. dpe 2keRex)""er (Mu \? (4) =") (84) dpe) _ 2ka(Rea)-"oc ( Mo \* (HF) ES (a) ee ‘Where pz, pc = liquid and vapor densities, lb ft” (kg m~),cazo HPING SYSTEMS ‘The folloving table can be used to determine ki, ke m, and nr Liquid | Vapor | Ren Ree ke ke n m t t >2,000 | >2,000 006 | 0c | 02 | 02 v t <1o00 | 32,000 | 16 006 | 10 | 02 t y >2,000 | <1,000 | oo | 16 02 | 10 v v <1o00 | <1,00 | 16 16 10 | 10 The value of 0.04 is based on smooth pipes, and ¢ and v denote turbulent or laminar flow respectively. +3. Calculate ¥ from te) /(#) “ x= |(@F oer (C8.6) (2/2), oo) 4, Use Fig. C85 to deiermine 4, or do. 5. Calculate the fiictioral pressure drop for two-phase flow irom either, dpe) _ gs (tee (22) = 3(%) This correlation typically allows prediction of the frictional pressure drop to about +3(%, but the discrepanzy can be greater uncer some circumstances. Where the flow is other than horizontal, the gravitational pressure drop be- cones significant, and in vertical flow it is the dominant term. It may be calcu- lated from dq (#2) = ssinélape + (1 er] (8.8) where g = aczeleration due to gravity, ft s-? (ms-*) @ = void fraction @= angle to horizontal, des Numerous correlatioas are available for the void fraction, and that of Lockhart ani Martineli is shown in Fig. C3.5. As it was derived originally for horizontal flow, it can be rather inaccurate fcr vertical flow. Other void fraction correlaions are mentioned in Chap. B.8. The accelerational pressure drop is often negligible for two-phaie flow of «ryo- genic fluids. I can be estimated frm, dD, Ld [Me ME (%) Adz [me pul = @), (9) where the expression on the righthand side is evaluated over the pipe section, ‘The total pressure drop calculsted by this method should be increased by at‘CRYOGENIC PIPING SYSTEMS Caz oot 2 04 at 1 100 FIGURE C85 Locthart Martineli correlaticn. Teast 30% for design safety. A final chect should also be made for critical, or choked, flow atthe downstream end ofthe pipe. An equation proposed by Sehwveppe and Foust? for the maximum possible flow ate is : 05 ae -ln~ (10) G), maxinum fiow rate of the mixture, Its“! (kg s~. speciic volume of the mixture based on homogenous flow, f Ib“ (a? 1) Me where M.cazz PIPING SYSTEMS sothat Mo 0a +Mu/ ou M+ Me (ca) and the partial derivative is evaluated under isentropic conditions. It can be calcu- lated for conditions at the downstream end of the pipe from the thermodyramic properties of the cryogenic fluid. An alternative Version, involving the enthalpy, f, which is sometimes nore cowenient, is os 2 7av av w) 4 (® ( op ] » (i) > where the two partial derivatives ae evaluated at constant enthalpy and constant pressure respectively. If M, is less than the assumed total flow rate, the aressure crop calculation should be regeated with a reduced total flow rate or an increased vipe diameter. M=- (Cz.12) OXYGEN GAS PIPING aE Fo: oxygen gis, the fluid velocity must also be considexed when determining the pipe size. Th: allowable velocity is a function-of the oxygen gas pressure and temperature «nd the material of the piping and its componenis. For systems with Pressures up ‘0 1000 psi (69 bar) and a maximum temperature of 200°F (366 K), carbon steel end stainles steel piping are acceptable provided thal the maximum allowable gas velocity as shown in ig. C8.6 is not exceeded. 09 1s0| AAT 200°F GAS TEMP. MAX) dl a a) INTERNAL PRESSURE, PSIG* FIGURE C8.6 Maximum veboity versus internal pressure for steel pipé= lines. (Ref. 3.)(CRYOGENICPIPING SYSTEMS e423 ‘When the velocity is exceeded, copper or a conper-base material'such as brass ‘or monel is required. The velocity criteria should also be considered at valves. For ‘exampk, sonic velocity can occur at throttling and safety valves. Copper-base materials are required for the parts of the valves where high velocity can occur. ‘When oxygen za8 impinges directly on ferrous piping, such as from a side back feed line, the allowable velocity must 2¢ reducec to one-half the values given in Fig. C86, or the impingement surface must bea copper-base alloy. When the velocity returns to accepteble levels, the copper bass alloy is extended for 8 diameters in pipe leagth before returning to ferrous piping. For a more thorough discussion of material requirercents for oxygen service, see Re’. 3 at the end of this section. PIPING: ARRANGEMENT Consideration ofthe cryogenic Auld ‘rapertes hay an effect onthe piping arrange- ment, Because the cryogenic fluid is cxlder than ambient ai, the continuous heat leak from ambient air to the piping system is a design consileration. ‘Whenever cryogenic liquid ean be trapped between two valves. a line block safety ralve must 2¢ provided to prevert overpressurization caused by heat transfer from the ambient air (o the cryogenic fluid, As noted earlier in the discussion of ‘cryogenic fluids, containment of the pressure increase caused by vapcrization of the tragped cryopenic liquid is not pra:tical and tse of a lin: block safety valves mandatory. Wher. eryogenic gas can bs tapped tetween'two valves, the preselre rise should be cacculated to ensure that the pipiag system design pressure is not cexceedéd when the trapped gas warms vp to ambient tempereturey Wher. the. design _pressure ean be exceeded, aline-block safaty:valre should te‘uséd. 91 es: ‘oTrags are:normally designed: into liquid piping systems > prevent indesirable heat leak from a:branch of the piping system When it is not-n use."To Jorm atrar at the branch and main. run connection; a.vertical rise equa! toa mirimum:of 2 pipe diameters is provided in the branch: Heat transfet into te nonflowwing-branch will caise the liquid to vaporize and a:gas-liquic interface will-form:in-the trap. ‘The. gis-liquid: inerface in the. vertical. rise will >revent'flow from:the main? run into the-nonflowing branch. Undesiratle, heat traasfer:and iquid loss is therefore seduced, When’ Lquid flow is require. through the branct, the liquit wil: flow through the trap: »» - ‘When a cryogenic ligui tin is initial putin service, the varm piping will case liquid flashoff, waich.could restrict th: flow during the two-phase flow transient period. When it is possible to. precool the lines; tae piping can be sized for liquic phase ow, which will result in smalle: piping. If rapid: cooldown is required, the piping must be sized for two-phase flow. ‘AS 2 good aprroximation, the amount of liquid required to cool down-aline is . We ~- €8.13) where W = liquid, Ib (kg) 4m = lineto be cooled, Ib (kg) Com = mean specific heat of linc, Btu/(Ib-F), (kt kg“'K-!) AT» = temperature change throvgh which ine is coolzd, °F(K) snthalpy of fluid as a vapor at ambient conditions, Btu/Tb (kJ kg!) initial enthalpy of liquid wvhen entering line, Bru/Ib (kJ kg") Hh)Gaza DIPING SYSTEMS REFERENCES 1. RW. Lockiart and R.C Martnell, Chem. En. Prog, 45,39, 1949, 2. JL. Schweppe and A.S Foust, Chen. Eng. Prog, 49, Symo. Ser. 5,77 1953. 3. CGA Pamphlet G-4.4, “industrial Practices for Gaseous Oxygen Transmission and Ditribu- tion Piping Systems,” Compressed Gas Association, Ine. Arlington, Va. PIPING SYSTEM DESIGN— MECHANICAL A.R, Belair Engineering Manager Praxair, Inc. APPLICABLE CODES AND MATERIALS ‘The applicable code for the design of cryogenic piping systems is tke ASME 31.3 Process Piping Code." H31.3 presents a design procedure to determine the stresses inthe pipe due to fluid pressure, ppe ard Ould weight, and therma expansion and contraction othe pipe. The code 2lso lists piping materials by ASTM specification number. The allowable stress as a function of design temperature is listed for each material The listed minimum design temperature is used to determine which miterials aresuitable fer cryogeni: service. Some of the meterias can be wed in lower-temperaure service, provided the specified impict test is performed to ensure adequate ducility at thelower tempera- ture, The B313 Code ako lists pipe fittings and bolting by ASTM number so that zmieterials compatible with the pipe can be selected. For more discusson of materials, see the section “Materials Used in Cryogenic Piping Systems.” Other piping design considerations such as corrosion resistance, chemical com- Paiibility, and material melting poiat can affect the mate‘ial selecticn. Some eam- ples of these sonditions include: 1. Copper-baie materiak arc not recommended for amnonia or acetylene service Ddecause copper alloys react with the fluid, 2, Stainless steel piping is recommended for flammable fluids such as hydrogen father thax lower-melting-point materials such as copper or aliminum ualess she piping is protected to preveat exposure to heat ir the event of a fire. 3. Copper-bate materials (such as monel) are recommended for high-velocity axy- yen service See Chap. 82 for discussion of formulas which can beused to cakulate the sipe wall thickness as required to contain the fluid pressure,CRYOGENICPTPING SYSTEMS 0.425 ECONOMIC PIPE SIZING When 4 piping system is being designed, an initial pipe size mast be selected before the piping configuration is developed. This subsection provides recommendations for selecting the initial pipe size. However, after the piping system is developed, the pipe size selestion is finalized by calculating :he pressure drop bated on the actual siping confguration and compaiing it to the pressure head available. For cryogenic liquid lines, the initial pipe size should be selected 0 that the liquid velocity is in the range of 5 ft/s ‘1.5 m/s). For cryogenic liquid lines, the initia. pipe size should be ielected besed on the available pressure head (sce section “Piping System Design—Fluids,” subsection “Pipe Sizing Criteria”). When estimating pipe tmgth and aumber of fittings to calculate the initial pipe size, add 50 percent to the total resistance coefficient, K, to allow for pipe loops and extra fittings as may be required ‘or therma: flexibility To ensire good process control, one-third 6f the total system pressure drop should be allotted to the control valves. Except for oxygen gas, as discussed in'the section “Piping Syster: Desi Fluids,” subsectica “Oxygen Gas Piping,” there ace no specific gas velccity limita- tions. However, the gas velocity is nornally less shan 400 {vs (130 m/s) to avoid noise groblems aad excessive pressure drop. For systems with pressures up to 400 psi (2750 KPa), gas velocitiesin the rang? of 50 to 2(0 ft/s (15 to 65 m/s) are common, ‘When the pipe size of the gas line affects the plant pover consunption, the initial pipe size can be selected by using the folowing equation. The equation calculates the pige diameter that provides the minimum tctal cost; that is, pipe capital cost plus aperating power cost. Econonical pipe sizing (Imperial units) D= coors, MEE ES (a4) EA Pa where M = molecular weight ~ T = absolite temperature, K U = average compression temp, K C = cost of power, S/kW-h Y = evaluation period, years E-= compressor overall efficiency A= pipe sost fraction, $ per 100-fi/in pipe diameter Q =flow -ate, ft'/h @ 70°F, 147 pia P = absolate pressure, psia D = inside pipe diameter, in PIPING COMPONENTS ‘Econonical pipe sizing (Metric units) mrucr)' ge Ea) Sex (cate D=(1:58)| ‘waere M = molecular weight T = absohite tempersture. K average compression - temp. K C = cost of power, $'megejoule evalustion period, years compressor overall efficiency A = pipe cost fraction, $ per 4vscimm pipedianeter flow mate, m/h @ O°C absolute pressure, KPa = inside pipe diameter, mm. ‘The types of valves used in cryogenic service are similar to those used ‘or conven: tional fluids, except that the valves may require unique design features due to theC426 PING SYSTEMS cryogenic fitids. Metallic and nonmetallic materials must: be suitable for the low= temperatureservice. See the discussions on materials in the section “Materials Used in Cryogenic Service.” Valve types used include gate, globe, butterfly, ball, creck, ard safety. ‘The stems of gate, globe, butterfly, and ball valves are extend so that the yalve packing and operator remain at ambient temperature when eryogenie fluid is in the valre body. The valves are oriented so that ‘he packing is ai a higher elevation then the valve body, so that a gas pocket cin form inside the velve’s extension tute, thereby making the extension tube more effective at insulating the pecking fromthe cold tenperatures, Figure C8.7 shows au extendedstem gate valve EMO. PARTNAME 6 SOMETEOLTING Boney easter PACKNO, ‘Your asians FIGURECE7 Extented stem gate valve, (Coutesy Prax- air, Ine.) ‘Yalves must be designed so that quid cannot be trapped in a porion of it wien the valves are cycled from open to closed. When a conventional gate valve is closed, it ispossible for liquid to be trapped in the bonnet area, If liquid is trapped, Feat input from the atmospheie will caus: the liquid to veporize and overpressurize the valve body. When a gate valve is ased'in cryogenic sewice, the bonnet ‘area is Yented to one side of the valve so that any vaporized gas sannot be Tapped inthe bonaet area. Figure C8.8 shows one method that is used to vent the bonnet wea of agate valve ‘With a conventional ball valve, liquid can be trapped between the ball and the(CRYOGENICPIPING SYSTEMS az? FIGURE C88 Gav valie—vasted bon. (Coury Pres, Ine! hag swrarast 2 ‘Seat. Similaily, for bail Valves' in’ cryogenic service, the ball area is vented to one side of the valve. .- _ Duting the’ cool-dovn' or thaw of & cryogenic piping system, different parts of ‘the Valies‘may’ccoY’ down, or Warm ‘ip at differen: rates, resilting in varying rates ‘of contraction aid expansion{’The valve design must consider that the valves are required to be operational curing thes» conditions. FLEXIBILITY ANALYSIS AND SUPPORTS FOR. CYROGENIC PIPING SYSTEMS Piping flexibility analysis’is ‘an: important design consideraton because the large’ difference: between ambient and cryogenic temperatures wil result in significant thermal contraction. ‘The:analysis methods used are similar to those required for conventional piping systems as discussed in Chay. B4. The one difference is that piping in cryogenic service contracts rather than expands asis the case with high- ‘temperature service. However, since the commerdally availzble flexibilty analysis compuer prograns have the temperattre range as an input value, and the program calculates the resulting thermal contradion, the analysis methods become identical to thos used for conventional piping «ystems. ‘When possible, cryogenic piping is iouted so that the pipng configuration pro- vides adequate fexibility. Generally, pping is routed in a manner that :ncludes Z, L, and U bend configurations to take up the pipe movernent while keeping the stresset within the allowable range. When additonal flexibility is required, the flexibilty can be increased by the addition of expansion loops (U bends) or the addition of flexible metal hose. Flexible hose can accommodate larger pipe move- ments and can result in a more compad. design. However, the use of flecible hosescaze >IPING SYSTEMS ay result in requiring more pipe supports to guide the. piping so that the pipe movement is taken up by the how. The flexible hose is tocated so that pipe movement + lateral tc it. The briided cever on the flexible Lose prevents any significant pise movement axial t> tho flexible hose. A flexible hose acts similarly to an expansion joint designed for lateral movements. The recom nended ugper pressure limit for 1exibie hose and expansion joints is 100€ psi (6900 kPa). ‘Hanging. Buh) ‘wi(mK) Urethane foam oor 021 Foanglass nz 042 Perlte (at atmogheric pressure) aos “031 Perlite (vacuum xt 1 pm) 19x 10" ee Laminar radiation shielding 21x 10" 383 x 107 (wieuum at 0.1 xm) The typical values for thermal ‘conductivity are shown in Table C8.12. The expanded foam insulations use a plastic covering, such as PVCor neoprene sheeting, to provide the vapor barrier protection. The initial sapital cos; is usually lower than the other systems, but more frequent maintenance is required to maintain a tight vapor barrier. Powder insulation is generally used vhen severil piping segments and picees of equipreent can be grouped in one area. A metel jacket or casing is used to contain the perlite around the piping and equipment When theinsulation space is maintained at atmospheric pressure, it must be purged with a dry gas, suck as nitrogen, to keep atmospheric moisure out of the casing For improved heat transfer performance, the casing cen be made vacuum-tight and the insuation spac: evacuated to a high vacuum. For powder insulation, high vacuum is a pressure level less than 1. um: of mercury when theline is at the normal cryogenic teriperature operation cxndition (1 yum of mercury is oqual to 10° torr). See Table C8.12 for typical thermal conductivity ‘values. When considering this insulation option, the difficulty in ma:ntaining’a ‘vacuum-tight casiag must be weighed against the difficulty of maintaining a dry gas purge, considering the difference in heit transfer performanze. ‘Vacuum-insulated piping (VIP) is constructed cf a stainless steel inner pipe that ‘contairs the cryogenic fluid, a stainless teel outer jacket to form the vacaum space, and insulation in the vacuim space. The insulaticn is normally laminar radiation shielding that consists of altcrnatc layers of a reflective material, such as aluminum foil, ard an insulation material, such is glass paper. For vecuum-insulated pipe, the recuired vacuum is a pressure level less than 0.1 um, when the line is at the normal cryogenic operating condition. Heat leak by conduction and radiation is reduced by the laminar radiation shielding, The heat leak by 2onvectionis reduced by the vacuum. Normetallic sfacers are required inthe vacuun space to support and maintain alignment of the nner pipe withia the outer pipe jacket. Bellows are required in one pize to account for the differential expansicn between the inner and outerca30 OIPING SYSTEMS yj Ly oman H MALE EW OF wer tt | | ‘BAYONET FIGURE C89 Vactum-insulated pipe bayonet assembly, (Courtesy Praxair, In.) TABLE C8.13 ‘Typical Heat Leak Values for Vacaum-Insulated Pipe when the Fluid ‘Temperature is ~320°F (-196°C) ee Baycnet and feld- Pige size welded yr Pipe Elbow Tee ~—Flex hose joits- Valve % 047 150 2.00 141 B 4 1 050 230 3.40 450 a 15 Ms 058 260 4.00 174 2B 19 2 064 = 3.00 450 192 a 25 3 07 3.80 570 237 a a 4 092 4.40 650 276 & 2 6 1.20 6.00 9.00 360 ww 120 8 146 740 11.00 438, 4 180 OO * (eat leak values are BTU/h/ft of pipe and flex hose and are BTU/h for each ftting and vahe) {For the inner pipeCRYOGENICPIPING S¥S"EMS 431 TABLE C8.13M (Metric) ‘Typical Heat Letk Values for Vacuum-Insulated Pipe when the Fluid Temperature is ~320°F (~196°C)* Bayonet and ficld- Pipe Size welded (am) Pipe Elbow - Tes Flac hose joints | Valve 19 os, 144 12 136 4 4 B 048 221 327 144 6 4 3B 056 2.50 35 167 7 6 50 062 288 433 18s 7 7 Bs 0.76 3.65 548. 128 4 9 100 088 423 635 165 2B 2B 150 Las S71 855 346 2 35 200 1.40 72 1050 421 41 3 * Hot leak valuesare watts/m of pipe and fx hose chd ae wats for etch fitting and valve.) + Forthe inter pipe. pipe. Due to the labor-intensive nature of fabricating VIP, it is'normally shop- fabricated. To accommodate field installation, a mechanical joint is required be- tween fipe segments. A bayonet assembly is one type of jointthat has ar extended heat leek path between the inner pipe and the flange connecton of the outer pipe. Figure 28.9 illustrates the bayonet assembly. ‘Table C8.12 shows typical values of insulation termal performance for laminar radiaticn shielding For a vacuum-insula‘ed piping :ystem, thethermal performance of the straight pire is affected by the umber: and type of spacers. The thermal performance of fitings and other components devend on fébrication techniques. Tables 08.13 and Table C8.13M (Metric) show typical heat leak performance values for commercially available vacuum instlated pipe. ‘The total cost, including the piping system anc the cost cf refrigeration losses due to heat leak, should be considerec when selecting an insulation ssstem. For many permanent piping systems, vacuum-insulated pipe prevides the ‘nost cost- effective insulation system. CLEANING * All maerials used in oxygen piping ststems or connected with oxygen systems should 3e cleaned before the system is put into service. The deaning must remove mill scale, rus, dir, weld slag, oil, grease, and othe: organic materials. Tke purpose ofthe deaning is remove hazardous hydrocarbans and pariculate costaminants which could ignite and cause a fire in aa oxygen atmosphere.To obtain additional information on cleaning requirements, see Ref. 2. For a discussion of the factors affecting ignition in an oxygen system, see Ref. 3. All materials used in nonoxygen cryogenic systems are also cleaned >efore the system is put in service. The purpose o: the cleaning is to reduce contaminants to the point where they will not migrate, seize up maving partset low temperatures, or prevent the attainment of high prodict purity or vacuum levels.e432 SIPING SYSTEMS VENTS SSeS When a cryogenic fiuic is vented. such as from safety valves, safe disposal of the finid must be considered. The fiuid should be directed so that it will not contact personnel, because the low temperature can cause burns. When cold, most cryogenic fitids are heavier than air. A heavier-than-ar gas can displace the air anc create a fotential for asphyxiation. Some cryogenic duids, such as hydrogen, are flammable An oxygen-enrichec atmosphere can promote fianmability of other materials. Therefore, all vents siould be Incated outioors ard directed so that high concentrations will not collec in confined areas. REFERENCES a 1. ASME B313, Process Piping, ASME Code for Pressure Piping, B31 An American National: ~ Standard, 1386 edition. 2. CGA Pamphlet G-4.1, Cleaning Equipment for Oxygen Service, Compresed Gas Asocia- tion, Arlingon, VA. 3, ASTM G88 Standard Guide for Designing Systems for Oxyyen Service, American Society “for.Testing and Materials, Philadelphia, PA 1984, i - : one COLD BOX PIPING 2 Je K. Howell, pate Be Cold Box Engineer Praxair, Inc.” A:zold box isa mechanizal system that insulates an entir: low-temperature process mechanical enbodimen: in a single assembly instead 0° insulating each pipe and vessel individually. The old equipment and piping are iastalled in an airtight steel insulation casing. The void space inside the casing is fillsd with insulation povder orfibers, +The design of cold tox piping requires knowledge of some basic engineering disciplines, such as strength of merials, fluid mechanics, and hea: transfer. Also required are some pipiag design skills such as flexibiliy analysis, pressure drop cakculations, and material selectior. These skills are discussed in Chap. B4 of this handbook. For safe design, colé box pipirg is usually designed to meet the engineering recuirements of ASME B31.3', Process Piping Coce. Further, the designer should beaware of tre design considerations for safe operation and mainenance of cold bores discussed in Sec. 8 of Ref. 2of this article. What follows is a description of problems unique t> the design of cold box piring, and, where appropriate, suggested methods of solution.CRYOGENIC >IPING SYS"EMS ~ 0.433, DESIGN REQUIREMENTS FOR COLD BOX PIPING Goneral System Architecture Cold boxes tend to be tall, vertical structures. They often contain tall distillation columrs and brazed aluminum heat exchangers which are oriented for vertical flow (see Fig, C8.10). A minimum of 12 in (3) cm) of srace is usually provided between the casing and the piping and equipment inside. This space provides adequate insulation for the cold equipment and piping and pipe supoorts must withstand oads from the insulation ma.erial. These loads occur when fibrous insslation is being packed into the casing and when powder insulation is teing loaded or removed. Reliability Piping inside the cold tox is not readily accessible for maintenance. Repair of a simple leak r-quires thet the cold box be out of service for several days. Eresion caused by insulation anc powder p:opelled by the leaking fluid canenlarge a mall leek or cause another leak in nearby equipment. For these reasons, leaktight integrity of cold box piping is extremely important. The number cf flanged joints indde a cold box is kept to a mitimum, Welded or brazed pipe joints are used wherever postible. Transition joints are often used to join stainless valves to alumi- nun pipe. Weld-end, top-entry, extended-stem globe valves afford maximum selia~ ~ilty. These valves do not have any flanges inside the casing, and the plug andseat can be replaced from outside the cold box. When fimges are ised inside the casing, prope: torquing of flange bclts must be assured. Threaded jcints and tubing, Compression fttings are usually no; used inside the casing. Flanges are sometmes insalled in a separate, small, insulated compartment where access can be achieved without removing the irsulation from the entire cold box. A careful leak test in Gore on all pping inside the casing before the insulation is installed. Thermal Expansion and Contra Typically, colé box piping is heated to 212°F (100°C) befcre start-upto remove any mosture, and then cooled to opera‘ing temperature—aslow 2s ~320°F (~195°C), as 's the case in an air separation plant. In that example, the total temperalure range is 532°F (295°C), The total thermal expansion and contraction of alumiaum pip? in thie renge is 6 ir/100 ft (0.5 cmim). The flexibilty analysis of whe piping must. consider the full temperature range as well as any more severe differeatial temperature conditions which may sccur during upset, ttaw, or cod-down, Clear- ance must be allowed for the pipss to move through their full expansion and contraction raage. This cearance is especially critical for pipes connected near the topof tall vessels. Pipe penetrations through the casing vill restrict the movement of the pipe unless the casing is made flexible by use of a metal bellows or a rubser boot Piping Installation Detzils Pipes running from liquid pipes or vessels to the casing must have thermal traps (upward rise of at least 2pipe diameters) to gas-bind the connectior and keep the Col¢ liquid away from th warm casing, Thermal traps are also used in gas pipes to prevent unvanted convection currents, Extended sem valves used in cold pipes are installed with the body located at eas: 12 in (30cm) inside the casing and with the packing glard and the operctor or handwheel ‘ocated outside, Liquid valve stems are tilted upward at least 15° to gas-bind the stem extension tube ard keep the liquid away from the packing. Supports forcoldpipesare commonly made from austinticstainles steel becauseCRYOGENIC PIPING SYSTEMS C435 it is song, ducile at cryogenic temperature, end has a -elatively low thermal, conductivity. Blocks of insulation material can be included inthe supports to reduc: heat leak from the casing. Low-point drains are provided where necessiry to remove proces: liquids for shutdown and toremove liquid water formed during thaw of the cold box. Cleaning Cold box piping is cleaned to remove contaminsnts (see the earlier dscussion of cleaning in this chapter). The cleaning is usually done before the pipe is installed Pipe ends and other openings are covered after completion of fabrication to main- tain clsantiness, DESIGN REQUIREMENTS FOR SPECIAL SITUATIONS —re Reversing Exchangers Reversing heat ecchangers and regenerators are often used in air separation cole boxes. The piping in the reversing streams is subjected to cycic pressure variations Fatigue analysis should be performed on this piging. Paragraph K3048 of Ref. 1 can be used as a guide for the fatigue enalysis. Liquid water may be present in the warm-snd piping from these heat exchangers. Fossible corrosion and freeze-ur problens should be considered in the design of tais piping: Liquid Oxygen Evaporation Most iquid oxygen contains some trices of hyirocarbon contaminants. If this liquid vaporizes in a dead-ended pipe o: crevice, tke hydrocaibons will concentrate, Explosive mixtures can result, Crevice: should be avoided ir liquid oxygen piping where a source of heat is present. Dreins and otter dead-eaded pipes connected 10 aliquid oxygensource should be trapped as close as possibe to the liquid source, Liquid oxygen piping should be well-iasulated from the casng and frem warmei process equipmert, Vacuum-Insulated Cold Boxes In sone cold boxes, a vacuum is pumped on the insulation casing to :eduice the heat leak. In suca cases, the piping ad vessels are tested with helium to find and repair tiny lerks which would greatly reduce the effectiveness of the vacuum, insulation. The design pressure for such piping is tased on the absolute rather than the gaige pressur. Instrument Lines Instrument pipes 1 the cold box must be designed with cars. These pipes have a small ciameter and are easily damaged. The design of insrument lines shouldC436 PIPING SYSTEMS provide adecuate flexibility, suppcrt, and protection fron insulatior and/or inedver- tent loads die to using these lines to support the weightot a worker during farica~ tion, Equation (16), paragraph 31).4.1 (c), Ref. can beused for assuring adequate flexibility ofinstrumen: lines. Flammable Fluid ‘Some cold boxes process flammable fluids such as hydregen, hydrcearbons, cerbon monoxide, and so fortr. Any aluminum or other low-nelting-tenperature riping outsidé the casing should be protected by fireproof insulction. The insulation should not contain any chemicals which would corrode or otkerwice degrade the piping material. Discharges from the dreins, vents, and relief valves must be piped to a flare stack orother safe neans of dsposal. Connections are provided on dead-ended ives for purging air from the lines before admitting the flammabk process fuids, REFERENCES 1. ASME BS13, “Frocess Piping,” Ax American National Stendarés, 1995 edition. 2. CGA Pamphlet C-8, “‘Safo Practices Guide for Air Separation Plants,”; Compressed Gas Association Arlington, VA, latest edition. LIQUID STORAGE AND CONVERSION SYSTEMS NH. White Engueering Associaie Praxair, Inc. Bulk liquid storage systems are often used at or near the point of find] consump:ion. ‘They typically supply product in three basic forms: as a gas at less tan the storage tank working pressure, a8 a gas at high pressure, and ai a liquid. The equipnent and piping required may vary, depending on which of these delivery forms is wsed, However, several general considerstions apply to the pining desigr in all cases. The working pressure of all lines in the system should be at least 10 percent higher than that of the tink. Relief valves required to protect the pping shouli be setat this pressure. This configuration of relief valve sei points wil ensure that if the tank is overpressured, it will be relieved in the gas phase fromits safety valve and not in th¢ liquid phase through one of the pipeline relief valves. Vaporizersare commonly used in storage and conversion systemsto gasify liquid preduct and bring it to ambient temperatures. The piping downsream of these ‘unis should te designed for liquid product temperature This should continue up to the point where a safely device 's installed to automatically shut off flow in the event that low temperamre is sensed. The response time of this device mus: beCRYOGENIC PIPING SYSTEMS cast compatible with the failure modes of the vaporization systen used. The set poin: chosen must consider the design temperature of all components downstream, Relief valves should be installed be:ween any set of valves that can trap liqui or low-temperature gas if they are simultaneously closed. Tkese blocked-line relie® valves are required to relieve pressure buildup due to heat atsorption. They should be set atthe line vorking pressure and szed to relieve gas at conditions of maximur, heat absorption. Particular attention should be paid to situations where the source of hea: is natura! convection associated with bare lines and ambient vaporizers ‘They ae of special concern because even moderate winds cin greatly increase the rate of heating and subsequent pressure buildup. Bulk liquid units afe typically émployed in intermittent flow sittations. For thié reason, liquid lines will frequently have to be coded down io liquid temperatures when product is demanded. Therefore, sich lines stould be sized to pass txe requirec flow in the gaseous state to avoid excessive pressire drop a: the start of demand Design for cryogenic temperature requires that perticular attention be 2aid to line flexibilty. Although large displacements are involved due to the extreme tempera: ture ctanges, they can be eccommodsted by conventional techniques, including fiexible hose sections and expansion loos: In additon, most piping materials experi: ‘ence reduced impact strength at eryogenic tempe-atures. Tko material ection it this chapter deals in depth with this topic: : ‘Thermal trapging is a piping techrique that can be used in any liquid line whether it is insulated or not; to mininize heat input to anidle piping section ot branch Where the branch is oriented vertically a gas-to-liquid interfice will be = established near the connection, and the majority of this secion will be filled with ~ gas approaching ambient temperature, “his configuration greatly reduces heat input + to the oranch whem it is idle, -” : LOW-PRESSURE BULK CONVERSION SYSTEMS ‘The most common bulk liquid units deivér gas ata pressure less than ‘he storage tank working presse. The working pressure of sich systems is typicalyy less than 250 psig (17.2 bat). However, systems vith 2s higa as 600 piig (41.4 bar) working pressure tankage have been used coumercially. The delivery presswe must be limited to approrimately 80 percent cf the tank working 2ressure to allow for tolerance on the safety relief valve seting, and ior buildup in produst pressure during periods of nonuse, Figure C8.1t illustrates a typical piping coafiguration, and item numbers in the following discussion refer to this fizure. ‘The lines cotinecting the vessel the to presse relief salety devices (No. 301 and No. 302) are sized to meet the reyuifements of ASME Boiler arel Pressure Vessel Code, Sec. VIII, Div. 1. Guidelines for the requiired salculations are given inthe ramphiet C3A S-1.3 “Pressure Relie( Device Standard:—Part 3—Stationary Storage Containers for Compressed Cases,” putlished by the Compressed Gas Associttion, Nornally both a relief valve and a burst disk am included, The relief valve is sized to handle the vapor generated due to the hect load resulting from the loss of insulaton (loss of vacuum). The bursting disc is designed to handle the simultaneous loss of vacuum and fire conditions. Where it is critical that the system not be aken out cf service for periodic inspection of the reliet devices, itis common to use 1 dual set of safety devices with a diverter valve. ‘The fll systems arranged so that the tank car be filled from either the top or the bottom. Tank pressure is controlled during the fill operation by adjusting theRiriteme@ a @ ®t sae | | (2uy ‘yoneag Jo ks2une9) woIshs worsoaues nq UNSSDEAO] PRO TUNE Le7bLIS PUO 9G6E90S $3UAL0d “s7} Aq PI43N03 OURaN youju09 aunssaud pur anzouaduas Ko7 « c.438CRYOGENIC PIPING SYSTEMS cam portion of refill liquid entering from the top telative to that entering from th: bottom with valves No. 402 and No. 4(4. Filling fiom the tog reduces tank pressure by condensing tie vapor in the gas space of the storage tank. Filling from th: bottom increases tank pressure because the liquid acts as a piston, comoressing th: {gas space. Special piping consideratiors for the fill ines invclve the line size, whica should be adequate to permit refill to occur in a reasonabe time period and yet not be oversized to limit heat input. Insulation is not typicelly used or these lines because of their short length, intermitent use, and high transfer rate. Helium ani hydrozen fll operations are an exception to this practice because of ther extremely low temperature and low heat of vaperization. Cther products may justify the use of instlation based on their value. Bulk tanks are normally filled fromtransports,which are discussed in the mobil: equipment section’ of this chapter. The transfer is ‘driven either-by a centrifuged pumpor by pressurizing the supply veisel. The ceatrifugal pumps are typically one, ‘or two-stage designs matched ‘for the discharge rate and pressure ty adjusting the speed and impeller cut as in conventional service. The 2ystenoctudig the the ‘transport piping, the off-loading pump or pressurzation cirent, the tanc fil pi and the. tank pressure ‘relief circuit—must' be. analyzed ‘te enstiré"that he tank ‘tank, canno’ be overpressurized in the refill process. Compressed Gas Asrociation positioa Papen S-8) “Protection of Storage. Tanks From Over-presure during, Operator Attented Refil,” addresses this topic ‘The tank instrument piping includes a gas phase'line called the full rycock that extends into the gas space to the full Iquid level. This line & used by tae operator during-manual-fil operations-to determine: wha the taik-is full. The operator opens valve No. 200, when the fil is vearly complete and observes when liquid is discharged; whic1 indicates the full level has been reached: Also inclided in the instrument piping is e Tiquid~ and gas-phase line used to measure contents: and pressure, ~ Control of pressure between fli accomplsted by the combined actions ofthe pressure building, economizer, and pressure-relief systems. The pressure-building circuit consists of components Nos. 500 through 503and is driven by th: difference in-demsities of the liquid and gas phases. It functions when product flows from the liquid phase line.is vaporized in the pressure-building vaporizet No.503,and returns to the gas space. This systemis required when the demand for product at the use points sufficient to cause the tank pressure ts fall. Attention must be paid to the very low driving pressure in-this circut. Preduet is nomnally delivered to the use point fromthe tank in liqud phase? Ia the case of low or intermittent demard, the proluct vapor pressure can increase above the delivery pressure. At this pcint pressure-building is no longe- necessary, and its desirableto withdraw product irom the tank in gas ptase. Product delivered in the gas phase will remove. approxinately five times the heat from the tank a3 the same mass w:thdrawn in liquid phase. A schene to autonatically switch to gai phase supply at Ligh tank pressure is called an economizer. A siphon gjcle econo- mizer is illustrated in items Nos. 550 through 542, together with the siphon line inside the tank. When tank pressure ii high, regulator No. 502 is closed and back pressure valve No, 550 is open. The siphon is broken in ‘his way, and gaseous produ: is withdrawn through valve Nc. 600. The siphon is reestablishec at reduced tank fressure wren the positions of valves Nos. 502 and 5{0 are reve:sed. There are alternative economizer designs, and care mustbe taken it selecting one because most are limited to low instantaneous flows, ‘The product isnext vaporized and stperheatedin the vaporizer No. 6(1. Vaporiz ers are classified zs ambient where the required heat is derived from the atmosphere,ca4o PIPING SYSTEMS eas powered where the heat is supplied from anoiher source such as steam, electricity, cr hot water. Relief valve No. 602 provides protection in the event of a blocked line condition, and special attention should be paid to the vaporizer power source in sclecting this valve. A low-tenpersture 2ipeline protection device is nornally included downs ream of the vaporizer to automatically shut down flow in the event of vaporizer feilure. Tris shutdown device is the transtion point between ayogenic and noneryegenic Pipe design. In addition, a pipeline regulator is usually used to provide coxstant iessure in the pipeline, ‘The low-temperature protection and pressure reguation module given as components Nos. 700 through 755 ha: emerged as the preferred way of providing these functions Regulator No. 703 is pressure-loaded sc that discharge pressure is agproximately equal to the loading pressure 15 read on gauge No, 753. Temperature pilot No. 750 opens and reduces the loading pressure when - ges temperature falls into the set control range, until No. 703 is completely closed, In some appications, gis is requited at gréater than tte working pressure of the tank, In such cases, it is pumpec in the liquid state, vaporized, superheated to anbient temperature, and regulat:d to the required pipeline pressure. A bank of high-presours receivers is usually included to auatch the demand 1) the pump dis- charge rate. The pump is typically a reciprocating singe-acting design ranging in coatiguration from simrlex to tripex. Figure C8.12 gives a typical flow diagram of a high-presstre conversion system. employing a standa:d cryogenic storage “ank. Item numbers referred to in the folowing discussion are given in this figure, Pping Practice for much of the system remains unchanged fom the low-pressure gas syslem just dkcussed. However, sp2cial considerations must be mace for the pping, both upstrean and downstream of the pump to ensure proper sysem operation, The fiuid at the sucion flange must be subcooled in order t) meet the net Positive suction head (NPSH) recuirements. The degree of subcooling required Tanges from 2 to 10 psi, depending on pump design, concition, and discharge requirements The liquid at the bcttom of the storage tank is normally subcooled byapproximately the sta:ic head of he liquid phase aboveit. The suction piping must minimize bot1 heat input and pressure drop to preserve the available subcocling. Pressure losses are primerily due tothe fluid acceleration -equired by the reciprocat. ing pump ani may be calculated using classical analysis. Heat nput generally increases with pipe diameter, but may be essentially eliminated by ‘he use of high performance nsulation. Insulation should always be used where the tank working pressure is low, the derzand is intermittent, or the pum discharge rate is low. It is wed almos: universally in hydrogen pumping systems The pump is usually contained in 2 small vacuuminsulated sump, which is comected to the storage tank gas and liquid phases, As the first step in prining the pump, components Nos. 600 through 807 are used tc fill the sunp with licuid The product vaporized :n cooling these components is retumed to the tank gas phise, increasing system. pressure and thereby available subcoolirg. The second step is to run the pump with valve No. 827 open, which circulates liquid thrcugh the pump and discharges it to the lank gas phase, cooling the pump compression chamber. Finally, valve No. 827 is closed, forcing the pump to discharge ageinst recziver pressire which is being held by check valves No. 830. Temperature seasor No 831 is used to check for pump prime.RU 42 oD aK cast(2uy ‘troxvig fo &seuno9) “sdund awu>toKis 10) wrshs awd LoydsoUNYL, E89 AINE NMISN31X3 WANA ANIOd ASN OL dna Tonaoad : \ NT ssn yoauwan / unaKnelSht eee vit [me “@) ee, sun i MMWR aD autya ouins aenuvadinat SIRWA LWA E cauwragy, gow 0s i; & SIV LNAWNSLSNE OL any ate Barn saw alae SAW, an3e 2za8ies yuazod ‘sp Aq paseno> AWA 30ND RR PAs ae O= || | NWA WOE WT : caseCRYOGENICPIPING SYSTEMS cass “Thepiping downstream of the pump nust be designed for cyogenic temperatures ‘and high pressures as well as the pulsations and vibration generated by te recipro- cating 2ump. Tenperature switch No, 837 is used to shut the pump off in the event cf low pipeline temperature. Although noneryogenic materials may be used downstream of ths sensor, itis commoa practice 10 use cryogenic mate-ials for all high-pressure piping in the bulk delive-y system. ‘Priming with this conventional pipirg arrangenent works well while the liquid level in the tank is high. In this case a circulation flow develops, with fluid geining heat inthe path through the suction line, the pumpsump, andthe réturn ine, which reduces its densit?. The lightened liquid flows up the return line and dumps into the tank. This circulation is driven by she ditferesce between the static'head due to the tank gas ard liquid phase conteats and that due to the less dense liquid in the return line. When the depth of the iquid contents is neatly equal tothe height’ of the cank, this circulation flow quickly cools th: pump and presents subcooled : liquid zt the suction valve to,enable a good pump prime. However, circulation flow | is reduced at lower tank levels bécause the static head of the tank contents is: less, | ‘and prming beccnies:impoisible: Figure C8.13 gives an improved arrangement | in whic" both’ the, suction ‘and return lines are connected to, the tank liquid | “phase through an extension of.thie vactum jacket. The two genctrationt- are 20p. rated ty a short baffle to promoié convection within the tak liquid phasé. The differeace in height between the traps within’ the extension provide: the head » needed to start the circulation flow. The heat picced up in traveling tc and from =the pump sump reduces the density o! the retured liquid, which mantains the circulaéon once siarted. This configuretion’ allows priming to occur regardless of =the tank level. BULK LIQUID DELIVERY SYSTEMS ‘ = Where liquid-phase product is‘fequirec’at the use point, refiigeration rather than the fluid itself is usually of interest. Therefore, minimizing teat addition through “the tank and piping is the primary design consiceration in these systems. Some. degree of pipe insulation can almost always be justfied. The tink insulaton perfor- mance is also critcal to satisfactory sytem operaion. ‘dedlly, the tank should be elevated so that theentire presure requirement can be provided by the static head of the tank conterts. This eliminates the need for pressure building in the storage tank, vhich can easily represent several times the heat load from all other.sources. A hosizontal tank is prefered to a vertical tank, in this case, both because the elevated foundation is less expensive. ard because the variation in static head, as the ccntents are withdrawn, is reduced. Where pressure building is required and the tank will not be elevated, a vertical tank is preferred co minimize heat transfer Urough the gas to liquid interfaxe area of the vesel. In spite of the éfforts, the fluid nomally arrives at the use point with at least a ‘small gas component, Normally, condifoning by « subcoole:, phase separator, of similarequipmentis required to achieve satisfactory process control. This hardware is located at the application and at a minimum delivers satur:ted liquid by venting the gas that is gererated in the piping :ystem. Figure C8.14 illustrates a self-contained system that delivers subcooled fluid. Tt consists a of heal exchanger formed by a tube within a tube, called an in-line subcoo'er because it replaces a section of the supply line. It is arranged so that the(roscers oN amd “S11 &q pasanog) “21 sexeag jo “xojooogns sjusBokm sunt PTX TNL NY. JOVIOLS WO INTAKE SFIOOIENS INIT NE |. og. u7uno a waziau so 7 27 NL “ZV HOSS 13WT annayd 21809 INGA moe g TSS an sara of 0 LxoanaL st anwa awa B. vassa a @ “3 2 wes 4 © same wt a mont c.44g(CRYOGENICPIPING SYSTEMS . cass main froduct stream flows through tke inner tube, and control valve No: 200, located at the discharge end, emits a small portion into th: annular space. This control valve is actuated by the differerce betweea the vapor pressure of bulb No. 256, located in the vent, and the refererce pressure set by regulator No.251. When the vapor pressur? of the vent fluid rises above the reference pressure, the control valve cpens. The reference pressure is selected t» ensure that cold gas—but not liquid—is emergiig from the vent. Th: temperature of the liquid in the annular space & reduced, since the pressure it maintained at a near-atmospheric level, Liquid flowing th-ough the inner tube s cocled ty the annular stream. MOBILE EQUIPMENT SYSTEM *. 5 R.C. Cipolla... sels en Consultant 02s 2 “Trucking Operations’ +. Praxair, Inc. oh Bulk aimospheric gases, oxygen, nitrogen, and argon are transported a3 liquids in varioue styles of double-walled tank trucks. This section presents the most common semitrler-svle transport vehicles wed for the distribution, of ‘amospheric ‘A doubled-waled cylindrical tank t-uck has a product tank, or inner pressure vessel, enclosed in,an outer casing or jacket. The inner pressare vessel is designed, ‘manufictured and tested to meet requrements 0! Sec. Vill of the ASME, Boiler and Pressure Vessel Code. The casing or jacket is designed t the requirements of CGA-241, Standacd for Insulated Cargo Tank Specification ‘or Cryogenic Liquid, ‘Compressed Gas Association, The annular space tetween the inner presure vessel and casing is insulated and evacuated inorder to minimize the amount cf heat leak from the casing shell to the cryogenic Iquid, Heat transferred to the product will result in a pressure rise, with the eventual release of product. The inner pressure vessel is normally supported or attached to the casing at or near the inner pressure vessel heads. The upper coupler king pin, landing gear support legs, aid tandentt axle assembly for the eryogenic semitniler are standard automotive components available to the trucking industry. Atmospheric: gases are typically transporte in low-ressure semitrailers equipped with an on-board pumping system used for off-loading the product to a customer storage tank. This pumper-styk semitrailer allows fer 2 maximem payload at a gross combinition weight limit of £0,000 Ibs (36,300 ke)—the road limit for a five-axe tractor-tailer combination. Inner pressure vessels are rated for maximum. allowable working pressure of from 40 0 60 psia 276 to 414kPa) and constructed from AISI type 204 stainless steel or 4A-S083, duminum alloy (ASTM B-209). Casings are constructed from either caon steel cr aluminun alloys. Carbon steel casings are used along with stainless inser pressure vessels, end aluminim casings ‘are matched to akiminum inner pressw'e vessels. A lightweight casing design can be achieved by providing closely spaced circumferential ring stiffeners. These ribs reinforce the cylindrical casing section against tle external pressure load which results from the evacuation of the insulation or aanular space.caas HPING SYSTEMS Atmospheric gases have different product densities, which willin turn reault in different-sized product-optimized semitrailers. When cptimized for an 80,00 Ib [35,300 kg] road limit, « nitrogen-ized semitrailer has ¢ volume of about 80(0 gal (0.3 m}); an oxygen semitrailer about 6000 gal (22.7 m)}; and an argon semitailer atout 4500 gal (17.1 m). Liquid oxygen, nitrcgen, and ergon are exempt fron the U.S. Department of ‘Tiansportation (DOT) hazardous material regulations ‘or design requiremeats of the semitrailer or package when the product pressure is maintaized at less than 40 psia (276 kPa] duriag transportation, However, DOT markinzs and shipping preparation, along with regulation: pertaining to the trucking-components such as braking and lighting systems, are all considerations wien designing a cryozenic transport. Basic requirenents for vessels, structural suprorts, strucural protection, and external piping are covered by CGA-341 specificatons. Piping systems for aryogenic transports can be separated into two categories: ‘imernal pipirg, which includes piping internal to the inner presswie vesse! as well as annular srace piping and exterzal piping, which includes all piping exterior to -the casing envelope. Both internal and external piping systems are normally ‘abri- cared with butt-welded joints,” INTERNAL PIPING Ax internal piping system is constructed from sfsinles steel for stainless: inner pressure vessels and of aluminum for aluminum inner pressure vessels.'A° typical internal piping arrangement bas the piping penetrating the vessel walls througa the réat casing Lead, as compared to a location on the ‘ylindrical-portion of the vessel.’ (See Fig, 8.15 for a’ typical piping ‘schematic.)' Piping sections inside the inner pressure vessel are not subject to an internal pressure loading. Design FIGURE C8.15 Schematic of typical intsnal piping arrangement.(CRYOGENICFIPING SYSTEMS c.aa7 considerations for piping supports and attachments, include loadings imposed. by road vbration and thermal contraction due to temperature as low as ~320.4°F (-1952°C) Antular space piping configurations must have 2onsideration given tc heat leak- age, flesibility, and trapping of liquid or gas. Since heat leakage to the cryogenic liquid it primarily by conduction down the pipe wall, the length of the piping run is of importance. Stresses induced by ttermal contraction of ‘he line or movement of the inner pressure, vessel relative to the casing are typicaly accommodated by providiag adequate.piping flexibility in the form of loops ani bends. Siress levels due to flexibility are generally ‘kept - within the allowable limits imposed by ASME. B31.3, Process Piping. Piping bellows to iccommodate thermal displacc~ ments; are, not generally.considered an acceptable way of providing flexibility to annular space piping. Road vibrations have proven to be detrimental to bellows- style fitings. rad Su ie ‘Testing ofintemal piping systems incudes pressre, leak, sad otber examinations when required by codes. Pressure-testirg procedures follow ASME, B313 ‘ments Nozzle petetrations at the iner pressure sesel wall may.requize exentina. tion of the welded joint using dye-penstrant techniques: Annular spacz_piping is thoroughly leak t:sted, since:any Jeaks would: deteriorate the: vacuum condition within the insulation space. A helium mass spedrometer technique is the most commen method aged for vessel and pping leak testing. EXTERNAL PIFING * mdssatnd toes “The exernal piping sysiem is commonly locsted side an operating conpartment atsthe tear of the semitrailer:-however. the pressure-building coil is located under the senitrailer forvard of the tandem axle assembly. The rear grouping his accident protecton featurs provided by the.tardem axe ubframe, ar bumper, and the piping compartments, epg Pe stainless steel external piping systems are used on all transports regardless of inner vessel construction, with the exception.of the pressure-building col. The coil isnormally constructed of aluminum extrided-fin tubing, which is flanged :o stainless steel inet and outet piping. The functicn of the cail will be discussed later. Valves, gauges, and other instruments and controls are not necessarily constricted from stainles steel materials: Bronzes, copper alloys; aid aluminum alloys are used for various component paris. Internal and external piping runs are joined atthe casing penetrations with piping flanges. For aluminum vessel designsthese flanges provide ‘simple: way of ccnnecting aluminum irternal lines to stainless steel external lines. As Fig, C3.15 schematically illustrates, eryogont eemitrailers typically have two liquid phase lines. The first piping run is used to fll the traier at the production plant and to unload the trailer using the off-loading pump. The second liquid line is used for the pressure-building syste. Pressure in the inner pressure vessel can. bbe increased by vaporizing a small amount of licuid. The ail with a large heat transfer surface can readily vaporize liquid and retarn the waned gas tothe ullage space of the traile:. This pressure buildup is perfommed prior 'o and during the off- loading process in order to maintain adequate suction pressure for the centrifugal transfer pump. Gas phase piping includes a pressure-reliet device line which directly communi- cates wth the vapor or gas space near the midpoin: of the topcenterline A spring-448 PIPING SYSTEMS joaded pressure relief valve and a Tupture disk device are normally provided on ihe cryogenic semitraler. These reliet devices are designed to maintain pressure at a safe level under emergency conditions, including exposure of the vessel to a fre. Pressure-rolief devices arc designed to the requirements of Section VII of the ASME Boiler & Pressure Code as well as CGA Pamphlet S1.2, “Pressure Relief Device Staadards—Cargo and Portable Tanks for Compressed Gases.” The same Piping run it also usedto providefor amanual vent system and a oad rellef circuit. The road relief valve, typically set at about 30 psia (207 kPa), controls the inner yessel pressare during ‘ransportaton. Venting of produst when the pressure reaches ‘he road relief valve setting can occur. A shutoff valve ‘upstream of the roac relief allows isolaion during off-loading and, therefore, permits the buildup of pressure ‘ an operating level tear the working pressure of the inner pressure vesse. A second gas-phase circuit has multiple uses as a gas-phase outlet, pressure- building coil return, and a transfer pump recirculation line, An optional sparger, or gas-ptase fill header, has various uses. A typical design ‘ns a pipe running along the top of the inner vessel vith a'series of holes which + 1 as spray nozzles for the: liquid, With this line tezd from tke main fil line, - transfer of iquid to the trailer cin occur in an all-liguid phase, all-gas phase, oF aay combimation. Gasphase filling of an empty trailer through the sparge: will cuuse a panial collapse of the’ gis pressure. This pressure-reduction by edlapse eiminates product losses that would occur by manually venting the pressure from te vessel. A sparger alo provides a means of uniformlt cooling down or shrinking awarm inne: pressure vessel. Uneven shrinking ofa inner vessel from near anbient tc liquid temperatures can cause terious structural damage to the vessel, sup2orts, oF piping, ces . Inner vessel pressure, transfer pump discharge pressure, and liquid contents are monitored with trailer-nounted geuges. A differential pressure indator is themast ccmmon device used fcr.contents measurement, Liquid and gas teps are provided for-this gauge. ‘ *Samplingof product n the innerpressure vessel is necessaryin order to detemine the Jevel of product purity. liquid tap typically teed extemally to a liguid line em be used ‘or sampling, ‘The liquid fill level can be detected by a small-diameter fixed-ength dip tube. ‘Flow of liquid through the line indicates the guid level it the tank. If the semitrailer is ased to transport various cryogenic liquids, dip tubes positioned at different levels ‘would be-previded. 4 o * ‘ As previoasly stated, the annular spece is evacuated. Pressure levels below 0.04 in (1 x 10“ metér) of nercury are achieved by pumping down the space through 2 Ine equipp:d with a sautoff valve. A filter on this pipe eliminate: problems with the insulatior material being carrisd to the valve or pump. Tho vecuuin love can be monitore¢ using the trailer-eqripped thermocouple auge tube. Vacuum level is determincé by a portable vacuum meter, Cryogenicpiping must be protected from overpressurizatici in any piping section which can be isolated. After liquid or gas has been transferred, the cold product warms up and builds pressure which can exceed the busting pressure of the pipe oo fittings. For this reason, block line safery devices arz a mandatory featine in cryogenic piping circuits ‘Thermal contraction of externa, pipes can cause high forces when constrained, ‘Typically, centrifugal transfer pumps and meter systems are protected from these forces by expansion bellows in the inlet and discharge piping. Piping bellows are made from ccrrugated s:ainless steel hose. -‘CRYOGENIC MIPING SYSTEMS e449 ULTRA-HIGH-PURITY PIPING Raymond Dale Woodward Manager Prazair Semiconductor Materials Praxair, In. Ulura-hgh-purity (UHP) piping systemsare found in gas aind liquid supply systems used inithe fabrication of electronic sem conductor chips and other related devices. ‘The requirement’Stor purity in UHP.sysiems ‘ypicelly involvecor ils of less than 1 part per billior.(ppb)-of foreign méxerials. © Semconductor fabrication| involves ‘oriming alternate laysts of very, of gemiconducting material on the surface of nonconducting miterial suchas silicon. ‘The' lines form-miniaturé ultfa-large-scale-integrated (ULSI) electronic sircuits‘on, a small silicon chip. Many small chips are simultineously fibricated n-a single’ silicon vafer. The line or feature size of ULSI chps approaches 3 qin (0.13 sm) or less. Any particle, one-tenth of the ine size can cause a defect in the circhit, resulting in chip. cestruction. To keep the percemiage outpur yield of defect-free ‘chips high, the gaces and liquids used in the fabrication of ULSI circuits mustbe -very. pure and free of particulate contamination. :*7-~>s9 0" ~The present requirements for ga’ purity are that trace gar contaminition must “be' held to. the parts-per-trillion (ppt) level. For garticulates, contamination with particles greater tan about one-tenth tke semiconductor feature size mist be held oh ‘Mos specialty gas supply systems ere from stecial highpressure, high-purity _ cylindess, or from bulk specialty gas supply systens if their wsages are nigh. Bulk specialty gas supply systeme replace the high prosairo eylinders typically uscd with larger volume containers such as high-pressure modules or tube trailers or special bulk containers for liquetied specialty gases. Specialty gas supply systems typically consist of the gas source (cylinder or bulk) with the appropriate gas cabsnet or gas panels vith contro’s to regulate the flow znd pressure to the cusiomer, valv» manifold boxes (VMBs) or panels to distribute tomultiple we points where appropriate; gas isolation boxes (GIBs) to isolate sourcesof gas at a point of use: and interconnecting piping. >iping maybe single-wall ordoutle-contained, depending on gas characteris- tics and local: codes and regulations. These installations. may include purifiers where eppropriate. 7 nat ‘A schematic representation of a typical supply system 0 a fab is shown in Fig. C816 ‘SYSTEM DESIGN AND 2 FABRICATION REQUIREMENTS. UHP ges supply systems must meet extremely stringent desipti fabrication, installa- tion, checkout, and commissioning requirements. A brief description of typical requirenents is prosented in this subsection. Design Requirements Designers of UHF ges systems need to consider many items often thought to be unrelated to the riping design. Obviously, the deigner must consider the basic system requiremerts such as flow and pressure. Basic materialcompatibiity within the system and especially with the gas itself must also be considered, since many of the gases are tcxic and/or corrosive and/or flanmable. But the designer must + also look at the entire facility design philosophy andbe familiar with the bisic safety considerations, codes, and regulations. What quantities of gaseswill be permitted on- site for use and for storage? How will hazardous gas monitoring and toxic gase452 PING SYSTEMS detection be handled? Answers to these questions may alter gas systems design, determine waich systems are to te double-contained, and how te containnent system is to be installed and moniiored. Once these considerations are zesolved, then the system desigrer can preceed and address such things as purity and contamination istues. Tradi‘ional industrial gas systems are considered clean when contamination levels meet parts-per-milion (ppm) requirements. In the semiconductor industry, however, contimination levels mist meet purity requirements on the order of less than | ppt. For instunce, particulate cleanliness may require the following specifications: «10 particles/f® > O1ym in ize (< 380 particles/m? 5 0.4 win in size) < I partich/f? > 0.. umin sie (< 38 particles/m? > 4 yin in size) ‘Trace gas impurities are typicelly tapped within purts of the systém during component manufacture and systen assembly, and later released into the process stream, Joining material, laps within joints, screw threzds, aad otier cavities can ‘trap moisture, welding gases, and other contaminants aid can become longerm ‘bLgassing sources, ofter referred to as virtual leaks, For this reason,such geometric discontinuities must be eliminated from the system by cesign and ‘hrough proper fatrication ard installation practices. Particulate contamination is found in many of the sazie places as the trace gas contaminants (ic.,.crevices, joints between mating pieces, dead-end legs off-the process strean). In addition, particulates become trapped within irdentations and small voids that exist within the natural roughness of mostmaterials, For this reson, all pipe and component surfaces exposed to process gas streams are: polished as smoothly'as possible to eliminate coatamination from thissource. Industry standards tyrically require surface finishes tobe electropolished toa value oféss than 7 pin Ra (0.28 wm Ra) roughness, Electropolishing also leaves an enriched chronium oxie layer or the surface. Alternative surface treatmentssuch as special passivetion techniques are sometimes used in selected system designs, especially for corrosves. Particles can also be generated by erosion of the prccess strean over surfaces of system components inthe flow pith. Also, contact betveen moving parts serapes offmetal particulates from contacting surfaces. Therefore, sharp ecges, rough sur- faces; and slicing or contacting surfaces must be climineted. from. tie design. Many process conditions promose the generation of particulates or promote the release of paniculates from component surfaces: High piocess gas velocities etode ‘Particulates from component surfaces. Vibration or meckanical shocks shake loose aiticulates resting on process flowsurfaces. These actioas can either be caused by external condtions suchas adjacen: rotating or reciprocating machinery that is not preperly dampened or isolated fromthe system, or from valve or repulator operation within the system. Valve or regulator operation may also generate particles resulting from contact and crosion of moving surfaces. Finally, transient process conditions cen also generate sontamination, Variations in gystem temperatures may enhance release of trace-gas impurities or cause themal exfansion and contraction that could cause particulate release. Outgassing and paiticulate shedding upcn system start-up can also release contamination into the process strean. The basic design conzept in UHP gas systems design is to design clean, build clean, and operate clean, The UHP gas systems should be designed ‘o be leak-tree, ourgas-free, deadspace-fiee, and particle-free. Componert selectionshould ineude a thorough evaluation of all the components in static dynamic, and impaxted corditions. .CRYOGENIC PIPING SYSTEMS cas Matetials of Construction ‘The opvious starting point in meeting requirements of this type is to se the right materals. All muterials must be compatible with the process gas itself in order to avoid such problems as corrosion. In :ddition, all materialsin the system must be competible with cach other to avoid chemical or electrical reactions between differ- eat miterials and again to prevent corrosion problems. ‘The only material generally acceptable for febrication of bulk gas piping and system componeats is 316L (low carban) stainless steel. In addition, the steel must have tow sulfur content in order to obtain better wold qualities. Where 0035 percem sulfur content is acceptable for normal applications, the sulfur conteat for UHP systets should be in the 0.003 to 0.017 percent range. Asthe semicenduetor industry continues to reduce feature size and line widths, component and tubing manufacturers have to evaluate and eddress ever-increasing concems over not just the sulfur content, which impacts both weldability of the sgstert and machinability of the component, but also such concerns as nonmetalli inclusions in the >ase steel. This has led to the increasing use of vacuum arc remek (VAR) or VIMIVAR and other specially produced 316L siainless stezls, Additionally, zas-absorbing materials (especialy nonmetillics) mustbe avoided: Contaninant gases dissolve into many naterials ofconstruction, This oceurs primar. ily:in aonmetaltic components such as valve seals and filte: cartridges, but it cas also oecur in the metalic parts. Since sc much steel exists in the system components, issolied gases within thes components could present a serious contamination problem. Special manufacturing procecures, as well as final deaning and condition. ing procedures, nust be considered t) reduce cissolved gases withir steels and other netals used in ultraclean systems. Many nonmetallic materials, partictlarly those used in the filter cartridges, tend to dissolve gases, primarily water vapor Thisis on¢ of the most serious contaminants in semiconductor fabrication. Special materials or special >rocedures to remove the contaminants are necessary 10 prevent these contaminants from entering the systert. ‘Any ‘material not used in the particular semiizonductor fabrication step could cause mmacceptatle contamination, Heavy-metal ‘ons cause severe problems when they eater the semiconductor fabrication process. Metals such as sodum used in the fabrication or cleaning process of ether metalic parts could leave undesirable metal ions on ot within the paris. These ions must also te eliminated as much 4 posible. Ideally, all pige and components for one pipirg run should be of asingle heat. This easures that chemical properties of the two >ieces at aweld joint are similar. In practice, this is often impossible. Mest important is to ensure that all tubing and ‘compcnents have similar chemistries to guarantee weldabiliy. If it is recessary to we a componen: with a dissimilar chemistry, care muist te exercised to ensure weldability and integrity. Staadard elastomeric seal materials such 2s Tefon®, Kel-F™, etc, are acceptable {or sels. However, if metallic seals can be desigred into the system, greater leak- tight systems are possible. For this reason, valves with diaphragm or bellows seal: cn the stems should be used to totally isolate the product stream from external con- taminetion. Purity and cleanliness are important, Outgassiag of dissolved or adsorbed gases must te reduced to very low levels. This is sometimes actomplished by specia. material processing such as passivating steels in acid mixtures or by subjecting steels :0 a double-melt process. It may be necessary to “bake out” some of theC464 PPING SYSTEMS components or the entire system, if the contamination is serious, 10 drive off the adzorbed gases. . Welding Conventional welding practices such as lap’ and socket welding ar2 unacceptable for UHP appication since they do not permit full penct-ation and leave overlaps ane crevices that are very difficult to clean, creating virtual leaks. Hand welcing, which leaves rough surfzces on the interior, is also unacceptable. Rough surfeces, as well as divoloration from the heat-affected zone, are sources of particulate generation. Rough surfaces also retain more moisture. All proces: line tube joining is done by autogenous orbital: butt welding 2er- formed by special automatic process welding machines. The joint t be welded is purged both irternally and externally with high-purity inert argon to evoid oxidation of the hot weld surface. A good orbital weld must be free of stain on-the Product sugface, can Lave no undercut or zoncavity of the Weld area, and must have a generally smcoth process surface inish. Weld buildup and:meta_puddling are Not acceptable. Weld samples (or coupons) must be inde ’at the beziniting and end of cach shift or whenever the weld machine- variables or parameters such as line size or heat Humber ire changed. These’ weld coupons miust be marked with the welder’s nate, and a log of all sample weld! must be maintained © 8 «= 3. = “"Weld requirements fcr field wells are the satié as for shop'wells. That is, all welis must be autogenoss orbital vclds wherever possitie. All welling and veld reparation taces place ir either a clzan room or a controlled work aiea to fmai:ain system cleariliness. Field welders mist pass the’same certfication requirements as, shop welders. Fabrication and Installetion Fabrication and installation procedures for UHP systems are not dramratically differ- ent from normal industril practice, 2xcept for cleanliness. Because cf such simiar- ity, sleanliness becomes «nalogous to quality or safety in that it involves a culture rather than a procedure. Frocedures are easy to establish; caltures arenot. A culture Teguires a completely new approact to the fabrication ard installaton procest, ‘Fabrication of UHP gis supply sjstems must be condusted ia as cean a mamer as possible. The work should be done in an enclosed shop where clean conditions can be controlled more easily. Generally, it is less expensive to buid as muct as, posible in the shop. To provide the cleanest conditions possible, ill component fabrication must be conducted in a clean room, preferably at least Cless 100 quality, as cetermined by federal standards, ‘The UEP gis systems must be fabricated as asystem, Ezch step in the fabrication and installation process involves UHP argon purges. Once the fabrivation process begins, the entre system should be kept under a UHP argon purge until the system is ready for introduction of the process gus at start-up. Since cleaninessis a prime requisement, the cleaning process is one of the most critizal steps in all stages of fabrication and assembly. As stated earlier, in addition to general clecning, cleauliness involves polishing (.e., riechenical, chemical, or elecropolishing). All par’s must be cleaned at critical steps in the manufacturing process and mist be Kept clean betveen fabrication steps and durirg shipping(CRYOGENICPIPING SYSTEMS case ‘The best method of cleaning is to elzctropolisk the component. This procedure actually removes some of the surface metal, along with any contaminants on the surface. All gas supply system components such as valves and regulators should be fabricated in a shop where cleanliness can be controlled, Alsc, some subassemblies and piping “spocls” should be fibricetod in a deanlincss~ontrolled shop area. However, fabrication of the gas distribution: piping to interconnect the various components is ustally done in the field zt the job site. This exposes the components and piping pieces to contamination from the atmosphere. Taerefore, the division of work between the shop and the fielc must be carefully considered, . Components must be packaged to ensure the cleanliness level is maintained: Double-wrappingin plastic bags is a minimum requirement. Piping must have the ends capped. The tubes must be pressusized with rure nitrogen or argon, or sealed ina pressurized container. Consideration must be given {o bota the final destination aid theroute of transit. Changes inaltitude, and hence presstre, could dariage pack Onc: the clean components arrive ai the plant ste, they aust'be stoied So, that the ‘cleanliness, of the components is maintained. While waitiag for installatidn; all cofiporents must be stored in their protective wrars, and only opened when ready. tobe finally instaled. Syd ws wee "ar cays rend Wp “ Spedal care aad procedures must te used to install’ the componéat Without introducing ‘unnec:ssary contamination. It may be nécessary''0 provide a portable clean room at the installation site to perforni some of the requized on-site fabrication in as clean an environment as possible. The clear room shuld be at’kast Class 100 quality. As mich of the subcomponent assembly as posible should be done: in this sortable cean room. After sutcomponert fabrication, all open ends..of “spool pieces” must be covered with an anti-static polyethylene or anti-static inylar wrap and kept seded until just prior to welding irto the fina assembly. Tt may also be necessary to construct clean environment around some system pacts to ensure tat minimum contamnation is introduced as'final component installation is performed, ©" ~~~ Sr To reduce the possibility of atmospheric contamination entering the partially. fabricated piping gyétem, the portion of the system that has teen completed must be continually purged with a pure, inert gas fitetec to submic-ometer levels. Since {inert gates are used for purging and for naintaining clean conditions, safety provis- ions must be made to protect personnel against gsahyxiation, Ne In addition, all ‘ools that might come in contact with process. gas surfaces inust be maintained in as clean a condition zs possible. These tocls must be specially cleaned; used only on the clean system; and stored in the clean room or a dean envi. ronment. ° Y QUALITY ASSURANCE : SREY ASSERANCR An independent quality assurance representative should be used to review and ‘approve all installa:ion procedures for the construction of a ga: supply system, and {0 condust cortain tists and inspections asagrced to by the owner. These procedures and tests should include the following activities: 1. Inspect the preceaned, UHP comporents prior to shipmert,C456 PING SYSTEMS 2, Review and approve written procedures for the fabrication and field installation of the UHP piping system. ‘Qualify pesonnel to work on the UHP piping system, including welder qualiica- tions 4. Approve weld coupons prior to actual production weding each day. 5, Observe and inspect a:tual const‘uction work to ensure that agreed-upon proce- dures are being folloved. 6. Witness leak tesis on the UHP piping system. 7. Conduct or witness analytical construction verification testiag ofthe gas system for moisture, oxygen, hydrocarbons, and particulates Juring construction. Toensure that minimal contamiration migrates into the UHP gassupply sysem uring ‘operation, the gas supply piping system must be as leak-tight as possble. Extensive heliam leak testing using either a helium-argor test mixture under pres- sure on the inside of the piping (outboard test) or pure helium on the outside with a yacuum on ‘he inside of the piping (inboard test) must be conducted in which edca component, piping system weld, and mechanical jpint is separately tested, Specification ofleak rates of less than 1.27 x 10°*ft*/hr (10*/sec) at one atmosplere ‘of kelium are typically required. Use of either the indoarc or outboard test metiod may be acceptable; each has specific benefits and problems. The designer must take {inte account vhich method is to be used. Constracticn verification testing of the system may bé cequired atseveral péints) in the project installation schedule in addition to certification at canpletion. The quality assurauce representative shduld slso oversee the ‘inal anialyss of the CHIP system to veriy that it méets all product purity specifica‘ions. CHECKOUT AND START-UP ‘The same care that was taken in the design and installaticn of the UHP gas sysem_ mutt be extended to sysiem start-up. If proper precautions are not taken, all the benefits of a UP installaiion can benullified with the intrcduction of gross cointami- nation into the system. Typically, tte gas supplier oversees the star-up of the gas supply apd disribution systems in coordination with the cistomer. Great care must be exercised so'as not to contaminzte the system. ‘The entire system must be systematically purged of the weld ergon and test ‘gases and replaced with product gas. For example, if an cutboard helium leak lest was conducted, it must be recognized that the helium has permeated into the elastomeric materials such as valve seats, and that trapped gases may be down- stream of isolation valves that were shut off to isolate segments of the sysem durng testing. Finally, since the output yield of semiconductor fabs & so dependent upon the purty and cleanliness level of the gas supplies, knowledge of sysem operation and cleanlines: level is necessary zt all times to warn of poor operation, bick- contamination, or imperding performance degradation. Continuous, automatic, ‘computer-controlled, analytical moritors have been developed to very rapidly sur- veyall critical operating parameters gas purities, and particulate cleenliness. These monitors analyze system maladies ard either sead out alarns with suggested cor-ec- tiors or correct system adjustments automatically.CHAPTER C9 REFRIGERATION SYSTEMS William Y. Richards Refrigerart Consultent Lincolnssire, ilinos ‘The bread term refrigeration refers to a general scence concerned with the use of produciag temperatures below normal for commercial or other useful purposes. Refrigeration piping is used in conjunction with refrigeration equipment. Refriger- nis arefluids whic1 absorb heat by evaporating at alower température ani pressure and transfer heat out when they condease at a higher presstre and temperature. ‘The incicasc in pressure necessary to elevate the temperature level is produced by compressor of the reciprocating, rotary, or centrifugal, type. In the case of an absorption system, the transfer of heat, thereby boiling the more volatile iefrigerant . out of asolution. : sien Peet yt Fluids wwhich-do not change state axe sometimes used:to transfer Feat in-an indirect system. Stch fluids. are: called secondary coolants. To:be classified .as.a secondary coolant, the fiuid must be used for the transfer of heat without a change in ts state. Brine, 1 solution of salt and water, is a secondary coolant. .. “ Man; fluids have been used as volatile refrigerants in the evaporation,compres- sion, condensing, and expansion cycle. This chapter will deal with application and structural design o: pipiag for the more commonly used volaile refrigerants such as ammonia and scme of the halogenated hydrocarbons. It wil also cover general ‘method: for other iefrigerants where specific tables are not prevented. Since volatile refrigerznts afeused in the liquid, vapcr; and misture phases, each-of these will be treated separately. : - Many fluids hate been used for brines. Origindlly, the tem brines epplied to salt solutions such as calcium chloride or sodium chiéride: The use of such salt Drines permitted the transfer of heatat lover temperature levele without introducing refrigerent of the volatile type into refrigerated spaces. These biines were commonly used for cold storage plants, ice plants, ot commerzial and process refrigeration. Solutions of glycols are also used as secondary coolants, 3thylene glycol and propylere glycol are mosi commonly usedfor this purpose. Several other eempounds or mixtures have been developed specifically for the purpose ofheat-iransier media, ‘These compounds are specifically designed to have high the:mal capadties, low viscosities, and other desirable properties for high heat transfer and low pressure losses, C457case PPING SYSTEMS ‘Two majar codes relate to refrgeration piping. One of'these ii the American Stndard Safety Code ‘or Mechanical Refrigeration. This code is reviewed and revised periodically. The most recent edition at the time of this wrting was issued in 1994. This code is sponsored by ‘he American Society of Heating, Refrigerating, and Air Conditioning Engineers ard has been adopted ty many states and munici- paities to be the existing law in these localities. This Code will be referred to frequently in this section and will >¢ designated as ANSI/ASHRAE 15,1 Another important Code on pping is the American Standard B31 Cod: for Pressure Piping ? Section B31 5 covers refrigeration piping and rela‘es to structural design rules, fabrication, construction, and testing. The code in this section wll be referred to as the ASME B31.5 Code. This section recognizes and use: the definitions included in Sec.3 of ASHRAE 15. The definitions included in See 500.2 of the ASME 831.5 Code are also used. Ingeneral, th: definitions in these two Codes coincide with defnitioas in the ASME, Baller Code, Sec. VIII, usually called the Pressure Vessel Code.’ ‘n addition, the ASME B31.5 Code recegnizes and refers to the basic definitions of the Amecican ‘Welding Society. Other basic definitions accepted by the refrigeration and ait-coniitioning indus- tay are given in ASHRAE Terminology of Heating, Ventilation, Air-Conditicning and Refrigerciion, published by the American Society of Heating Refrigersting, and Air Conditioning Engincers.* : REFRIGERATION CYCLES Compression ‘System: Figure C9.1asbows a typical single stage refrigeration cycle plottec on a presture- enthalpy chact. Figure 09.16 is a typical diegram of a single-stage compression system and shows a compressor;.a condenser, an optioral receiver, an. exparsion device, and an evaporator. The state points on the line diagram of Fig. C9.16 are nunbered to correspond to the same. points on the pressure-enhhalpy chat of Fig, CO.1a oa In a typical system, F, represents the pressure in.the 2vaporator corresponding tothe temperature at which the réfrigerant is evaporatiag. From foint-1 to point ‘2yihe refrigerant vapor & cartied tc the compressor in a suction line, The presiure- enthalpy chant indicates a small pressure drop in.this line. From foint.2 to point 3, the refrigerant vapor is compressed: The connectior, between points 3 aad 4 represents the discharge, or hot-gas, line, and. the_pressure-enthilpy chart also incicates the pressure drop due to frctionin this line. Destperheating and condensa- tioa of the reirigerant ai constant pressure in the condeaser occur 2etween points 4 and 5. The iquid line s represented by the section between poin: 5 and point 6. From point ¢ fo point 7, there is a representation of the pressue reduction oF “expansion” through the expensior device. From point Sto point 8, the refrigerant is < mixture cf liquid and vapor because of the expansion at constast enthalpy. At point &, the liquid and vapor mixture enters the evaporaior; beat tmneferred ‘rom the evaporatcr results in evaporation of the liquid ia the liquid-vapor mixture and, as indicated, a slight superheating of the resultant vapor. At pont 1, the cycle is completed In this chapter, the line between points 1 and 2 will be referred to as the suction line, The pipe or tubing between points 3 and 4 will be designatec as the ho-gasREFRIGERATION SYSTEMS PIPING C459 Pressure Exponidn device i oy FIGURE C91 Compression refrigeration eych. (Courtesy Carier Air Conditioning Co) :caso PIPING SYSTEMS: lise. The pising between points 5 and 6 will be termed the liquid line before expansion, and that between points 7 and 8 will be called the liquid line after expansion. Ascorption System Figure C9.2 shows a typical piping arrangement for an absorption-refrigeration cycle in which a lithium: salt solution is used with water as the ref-igerant. In such 3 ‘ concen sever Cohan Cee Worm achution s 1H] ect wong ii | 2 de mF A wang | | | <> Semen on Cool soliton & ¥ <0) cycle lithium bromide type. (Ceurtesy Car- FIGUREC92 Absorption refrigsrat rier Air Conditioning Co.) fe ystem, the absorber and generawr serve the same purpose as the compressor in a compression system, “he refrigerant is absorbed in solution at ow pressure in the absorber and is pumped to the generator, where th: refrigerant is boilec out of the soluticn at high pressure. The state points are aumbered in Fig. C92 to corespond te the similar poiats of the compression sysem cycle ia Fig. C9.la, For other types of atsorption systems (such as ammonia and witer), a recifier ora fractionaing-colum type of parifier would be installed betweca the generator and condenser. Also, the evaporator might be remote o: of some cther type, such as a flooded cooller ‘or a direct expansion type in whh the refrigerant passes through coils.REFRIGERATION SYSTEMS PIPING cast In ¢ consideration of the piping of an absorpiion system, the pipitg between points 3 and 4, 5 and 8, and 1 and 2 may be handed in a manner similar to those of a compression system when the different fluids and flow rates whica might be encountered are taken into account. The solution lines between the absorber and generator and through the heat exchatger would be treatec as brine lines in the case of lithium bromide or lithium chicride systenis. For ammonia-water systems, properties of aqu‘ous solutions of ammonia are available, end these ines would be desinged in the same manner as would be brire or water lines, Present-day atsorption-refrigeratior systems are integrally designed as com- plete units, and in many instances th: enitre assembly including the piping is made at the factory. For these reasons it is seldem necessary to consiler piping for an. absorption system as.a.separate design prablem. However, the principles involved, in, this chapter. could :be: aprlied.to th: corresponding sections, of an dk eLaat ‘ton: (KW) of rafrigeratioa or in Ua gel git ot Btu: per hour: flow Fate Of the ‘rrigerint.'Sinée' the refvigerans'chaages state inthe: procé’t it's customary, to caletlate refrigerant flow in poulids per hour er per minite (ky/8) ‘This Weight flow ralé fora constant raté of reftigéraion will beconstant ‘throiugiiout the system (for a single-stage'system). The volutie of refrigerant handied at the arious doints of the cycle caif be Ueteriined ifthe density at these various points is known: Sinice the Voliimé “handled in‘ suction lie is important not orly ‘for the Piping selection but also for the compressor sélection, charts are'frequeatly made Sowing the cubic leet por shinute per ton (M/s HWY) of refigerant gat of vapor at this point. ~ : pee eBid “ To-calculate the flow rate’ of refrige-ant for a given’ rate ‘of refrigeration’ the following procedure is used:' : at Ib 200 E Gino iD en Gog ye (4 Way, 7) (219 where A, is the entialpy of dry saturated vapor at the evaponttor outlet pressure or temperature and hy is the enthalpy of the liquid refrigeraat at the expansion device irlet. cee ‘To calculate the flow rate it is necessiry to have thermodynamic properties of the refrigerant available. After the mass low rate has been ditermined 28 shown in the equation, itis possible to determine the volune flowing at various points in the system. ma 4462 PING SYSTEMS If the specific volume of the refrigerant is known at the state point leaving the evaporator corresponding to point in Fig. C9.La, the volume at that point can be deiermined as follows: cfm 200V or Ge =p (9.2) M - oY KW ~ Al (C2M) where V is the specific volume of tke vapor at the evaporator sucticn temperature. Fo: purposes of design. this value may be taken as the specific volume of dry, satarated vapor. In the previous equations, A, corresponds to the erthalpy of the refrigerant vapor at point 1 leaving the evapcrator. In the flooded type of cooler, this state point is close to saturation, but for a directexpansion coil-type evaporator, the vapor may beslightly superheated. For low amounts of superkeat, i is permistible to gnore the «uperheat for purpose: of selecting the piping, although this should not bbe done for compressor telection or for very accurate determination of the velacity. The value hy corresponds to the enthalpy of the liqud entering the expansion device. Some subcooling of liquid may exist at this point, in whici case poiat 6° woald result in a more accurate igure. However, for selection of piping, it is customary to use the enthalpy of the saturated liquid at the condensing temperature (pcint 5) for determining the flow rate. Later considerations will show when the actual state of the liquidmay have an effect on the liquidline size. A later exanple wil illustrate methods 0 calculation for flow rate by weight and by volume ix the suction line for a typical rofrigeran!. Table C9.1 shows flow rates in pounds per minute per ton and cubic feet per minute per ton at the stction condtion for three common refrigerants: ammonia, refrigerant Ri3da (tetrafluoroethme) and refrigerant 22 {monoshlorodifivoro- methane). This table is based on enthalpies read at saturated conditions of refriger- ant vapor anc liquid as previously mentioned. The cubic feet per minute per ton (dm? per second rer kW) ir the discharge line cannot be calcalatec readily because the actual temperature atthe end o/ the ‘compression a function of the compressor design and efficiency and these will vary among various manufacturers The discharge cubic feet per minute per ton (dn* per second per kW) for single-stage systems can be approximated by the following fornula: + P Discharge cfavton = suction cfm/ton x zB x12 (@3) ~% », (Discharge dm? e- KW = suction dm®/s-KW x axl 2) (93M) where P, = absolute pressure at sustion Py = absolute pressure at discharge This formuiais not exect, but it will serve as an approximation wher it snecessary to determine approximace velocities in discharge lines. -REFRIGERATION SYSTEME PIPING rnd TABLEC9.1 Refrigerant Flow Rate > . Rida Refigerant RI CHRFCS po. chemical formula NHS AeA ane | (common nares) ammonia Teuetiorocthane | “Corcitwormetnane Condensing toraperatice os StS 6 8 a Evaporatng temperatire “c © ° 3 10 4 20 -4 ano oats 0.420 “2 0422 0.00 aed 09 ose 04 Ca 1086 les/mic per won* 0429-28 S02 «AT 2.70 044 300 be 276 0440 312 528 345 2K s 251 287 295 oar 325 saa 363-291: 304 104 292 3.00 308 317 Cimiton (susion line)F 0 32 195 20) 2043438336207 280227 -10 16 283 2 29 SAL 538 297309 3.22 20 sé 431 445172114450 4.09, 30 22 668 6817001201 691343 650677708 SS seer *To convert to kg per kw, mubiply by 2.1494 x 10° To convert to dims per kw. multiply by 0.132 ALTERNATIVE REFRIGERANTS Because of the depletion of stratosphericozone, which is partly attributed to chlorine release¢ to the atmosphere from chloro-iuro-carben (CFC) chemicals, many alter- native refrigerants made from fluorinated hydrocarbon gases are being cvaluated within the refrigeration industry. The common CFC refrigerants R-12 and R-502 are no Ionger produced. The hydro-chloro-fluro-ca:bon (HCFC) refrigerants, such as R-22, are to be phased out of production in the United States by the year 2015, and earier phaseouts are scheduled in Europe. ‘The satural refiigerants ammonia (R-717),propine (R-290}, and carbon dioxide (R-744) are being promoted in new apstications because when released they do ot contribute to ozone depletion orto global warming, In addition to the customary use of ammonia in industrial food processing and storage, armonia is wow being widely used in district cooling plants fer cooling of commercial buildings. Some food processing plants, now using R-12, 2-22, or R-502, are converting or planning. to eventually convert to ammonia. Some new processing plants are being designed to use R-22 currently and to convert tc ammonia in the futire. Propae is now being ued in domestic refrigerators in Burope ani Asia. Carbon dioxide is now being used as a lov-stage refrigerant in supermarkets. - In the United Siates, the Environmental Protection Agency evaluatesand regu- lates outstitutes for the CFC ozone-depleting chemicals. The Significant New Alter- natives Policy (SNAP) rule lists substitutes for R-11,R-12, and R-502in both retrofit and new equipmert.case IIPING SYSTEMS * For new centrifugal chillers whith previously used R-11, R-12, R-113, R-1i4, or R-500, the substitute refrigerasts are R423, R-124, R-22, R-134a, R-DTea, R-717, anc absorption systems. * For new reciprocating chillers, which previously used R-12, the cubstitutes are R-22, R-134a, and R-227ea. * For direct sir-conditicning applivations, previously using R-12, the popular sibsti- tute is R-134a, * For cold storage and retail food refrigeration formerly using R-12 and R-S0?, the substitutes are R-22, R-134a, F-227ea, R-402A, R-02B, R-404A, R-S07, and R-717 vapor compression. TABLE C9.2a ‘Temperature-Pressure Chart for Constant Boling Point Fefrigerants nm? | RuBe Rso7 | 4290 | R744 sia pola | asia | sia | Tempe oaz 1069” 3933, ~30 Oss. 1502197887 -2 112-2965 438532 1738-420) 5704 4.856 230 54S TSM. 6180 - . 929 73H. 91899.» 8200... 1047, 426 93 1276 a ‘0.03 o82° M6 BIG LOR 9 749 1526. 4210S 40 982 1936-541 BLAS. “3 9 3% Twa. 2 154529233816, SO ay “3 39253558 a6” Bass 2657 “4 BN 299° 553029323 5 2908 sias 5129 OT BHD “ 33000 2 Ts 9 4B B a4 2 91 BI SSS 2 SO BAIT hagas BBM 58 a fo ~~ 8890-6012 9880. 34. BaP” 630 50 4S. 10836 7088 159 BS AGA vat 3 2°20 SARAOL, 8290, 12D BS IAS BOB @ 2S 145097965256 1868 BA 932, ” 300 16880. , 9 2Bs IBS 1069 % 3819535” 1286 1967 OB THT 95 40° 22498 a7 ms Ra ms 4s 2578016821509 5ST MS 13 so Saas TG m 55 Ben | 6s ISS BB HE. BL oo 57823 as 520 TaD 1» 6 645 7413763 10 7 79.12. 30691347 3B 38 0 spss 3819313 asia vs 50 nem 40S sa83 569 By 10 yous. 5761 a2 M0 ness 20 Pressure is ir psi 10133 for gouge presse), tabtrad 14,696 for gauge pressure (for KPa, maltply psa tines 68947, abiract[REFRIGERATION SYSTEMS PIPING C465 TABLEC9.2b ‘Tenperature-Pressure Chart for Zetotopic Mixtums with Temperature Glider . 118 0408 nom 196-1379 3.16 1632 =10165 °° 10136. 7738 sia 6427 3990 | 4137 <1) 55.15 81085516 e425 =7399 “89S 95 3225 35.84 -22 653 -28 26.43 333 Ad 3465 1835 sasm- 1517 al “un 39131193 a3 567 3330 1 268 1397 -023 ag | 244 1088 418-238-2325 200 -6y 932-1897, 182096 183 1358-1487" 1473, 32 225 HS YTSS 1108. 1080 a7 as 2652-240 ~ 224-4037 1st saat 520 536 a 2638 4147-1200 1217. 5816 Be 47891818364 | ons 3958, 5379 BRS 2404, ows 0st 635° 440334? ana oa 7365 0993199659 6a ar99 sia” 5123" s303 1673 sos7' $833 3844 - 2a 83.98 9653 6508552613 so72 y0298 zt asa? 97.01 10899 T7048., -77.30-.. 1655, 122 n4ee 8253 Ly 8275-1793 108.48 11995 767” 87.89. 19 11878 9757199739205 vat 13x63 1088010601248 13667 6S BTS 58 1462 tse0o 1277, z9s 5108 15525 i64ai 15108 51.22 3447 1388 13868 REZ BT 17900 4s us.83 41ST 1a504 152221229 4481 For guuge pressure n psig, subtract 14696: For auge pressur in KPa, subtract 10133, Fortemperature in"C = 99(7F-32) , ‘The-narketplaceis determining the popular subsitutes for tie CFCs. The HCFCs such as R-22, R-4)2A, and R-402B wil be evenxually phased out. The present choices being made in the United States, as reported in the commercial news, are: + R.134a and R-22 to replace R-12 in refrigeration systems * R-3la, R-407C, and R-410A to replece R-22 in air-conditioning equipment * R-404A, R-407A, R-S07, and R-717 to replace R-22 and R-S02 in refrigeration systenscase PIPING SYSTEMS Tables C3.2a and C9.2b show the temperature-pressure relation for the main czndidates. ‘The pressures shown may be expected in refrigerant lines wher both liquid and vapor are present, Propane (R-290) is incuded, hovever flammable options are not used in equipment exposed to the pablic in the United States because of lability laws, Table C2.2b shows the pressure-temperature relation for biends of two or more componeats. These mixtures lave temperature changes, called temperature glides, during viporizing and condensing. Note that R-404\ and R410A have very litte temperature glide, so that they may be created similarly toazeotropes (mixtures with a corstant boiling temperaure, such 1s R-507). These ard similar mixtures vith. very small glides are infornally referred to as near- amotropic refrigerant mixtures (NARMS). R-407A, R-407C, and many other tlends hhzve significant temperature glide. With careful design of evaporators and condens- cn, the glide can be used to improve efficiency. However, blends with significant temperature glide must be charged into the system in the liquid phase in order to maintain thepercentage of the componenis. Also, blends with glide are not sutable for use in flepded evaporators. PRESSURE DROP IN REFRIGERATION PIPING ‘There are tvo separate tasks when designing a refrigeration piping system The first task is to select the size of a pipe for a given flow. The second tsk, determining pressure drop, is accomplished after the pipe size is serected. Boh tasks use the same“equaticn, which describes the pressure, drop in fitid flow within piping ‘However, the tasks must be adcressed separately. Customary piping tables, such asC9.4 through C9.9, lst capacities at specified temperature or pressure drops per unit length cf piping. ‘These tables arc an aid to detemmining the pressure drop after the equivalent lenzth is known. But the equivalen: length depends upoa the diameter selected. Thusit is clear that the listed capacities are only correct for the unit length basis used in developing the table and are not to be used as reom- mended capacities. Daroy-Welstach Equation The pressure drop of most fluids in. piping is described by the Darcy-Weisbach equation: L vy? =f (£\(.% 9.4) a= (S) rx) eA (=(5)2)) an where Ap = pressure, pel (pa) ‘f= Moody fricion factor, dimensionless L = length of pipe, ft (m) D = inside diameter of pipe, ft (m) censity, 1b/2 (kg/m)REFRIGERATON SYSTENS PIPING caer V ~avelccity, ft/s (m/s) : envi is = gravity acceleration, 32.174 fs (9.807 m/s!) oro Optinum Sizing ’ In order to selec: a pipe size, some previous experience or citeria is needed, Sther than the customary capacity tables, Figire C).3 illustrates theeconomic:of selecting ; eet of ud rictién andifsttled pice ® 5160 ten Pie; 8 Cost per equivalent fot 3 ° 2 4 6 8 10 Nominal pipe diameter. (IPS) ‘A. Presert value of fluid fricion cost, : B. Cost olinstalled pipe. . C Total east . FIGURE C93. Cost of uid friction and initalled pipe. : 5 « pipe size. Initially, it may seem that for a giver flow and refrigerant condition the opimum dianeter of a long pipe will be greater than that of a short one Richaris’ shows, however, that the optimum diameter is irdependent of length ‘The method sets the present installed cost of pipe minus the present worth of the lifetime pumping costs equal to zero, and then differentiates with respect to diameter to determine the optimum diameter. The lengtk cancels ia this analysis, whichcasa PING SYSTEMS proves that the optimum diameter is independent of length. The effect of longer lengths is to increase the power cost for operation and may, for extreme Jergths, require a slightly larger pump or compressor. The first cost of the rext-larger pipe size will always exceed the savings represented by the present worthof future power costs. The tozal owning costs of a smaller pipe size will always be greater than for an optimum choice, because the present worth of the feture cost cf operation will always exceed the first cost savings. Studies'? recently made for R-717 and R-22 give some guidance for recom- mended piping capacities. Conservative conditions are assumed in ammonia indus- tial applications. Assume typical equipment efficiencies, running 4) percent cf the time at full load for 15 years, with power cost of $0.05 per kW, the cost of money 5 percent higier than ination, and with installed insulated pipe cotts of $13.0) per inf dia per t of length ($1.68 permm of dia per m of length). The results ofsuch stidies can te expressed in reconmended velocity pressures per velocity head, Aplk = pV'Ng (Ap/ke = p V'I2). A more convenient form is to sate the reom- mended equivalent temoerature drop in a globe valve. Recommenced Temperature Drop Gradients for Vapor Flow in Industrial Applications © For R-717 vapor, the result of such studies suggests that the temperature drop jn a globe valve at lew temperetures should be aboat 0.25°F (.14°C), that is at —40°F (40°C), and at high temperatures about (.125°F (0.07°C), ie. 2 at AVF (25 + For R-22 vipor, the suggested temperature drop in a globe valve at low tempera- ures, shoul be about 0.5°F (0.28°C), and at high temperatures about 0.25°F (14°C), Practical temperature drop gradients can be much greater with higher installation and operating costs, with less running time, and with shorter capital recovery life. For each dousling of any combination of increased costi or reduced running time or life expemancy, the optimum capacity increases by.28 percent, the presure gredient increases by 5€ percent in copper tube and by 62 percert in steel pipe. ‘The natural refrigerants R-290 and R-744 use similar pressure gracients as R717. AI halocarben refrigerants use similar pressure gradients to R-22. Comparing Pipe Flow Capacitios ‘To compare pipe flow capacities ve can rewrite the Daréy equation in terms of mass flow. Since mass flow equals area times velocity tines density, we can insert the mass flow divided by area and density in place of velocity. Ap = 15) ( we _\ (@s5) Dei epea) L we) (e»=1(5) (=i) onoREFRIGERATION SYSTEMSPIPING caso where ¥ A flow, Inms (kg/s) inside flow area of pipe, ft* ‘m*) Reacranging the equation to show the mass flow capacity, _ (Zest dpeprD) we a( Fad (C96) 2aAp mp D\% fet Now, since it can be shown that all pipe sizes have similar L/D ratio: and since the pressure drop and density are relatively cons:ant for a 3iven application, we can show that the mass flow is proportional to the area divided by the square root of the friction facior. w= al (96M) we 4 9 ‘The relative values of fricion factors for fully turbulent flow extend ‘easqnably ‘well in the area ofthe Moody chart between laminar and fullyturbulent low where most applications occur. Table C9.3 cempares optimum pige flow capacities for R-717 suction lines compared to capacites listed ir: Table C9.7. The friction factors used are those for fully turbulent flow. ‘The obvious point 1o be made in the previous comparison is that the listed capacities in Tabks C34 through C9.¢ are not recommended capacities. These tables merely provide a convenient starting point for adjusiing the cagacity and tempersture drop to other lengths and to other capacities. After a pipe size is ‘TABLE 59.3 Ammonia Vapor Flow Methods Compared PS rfim | 2] 3} «| s} 6, s| 0] 2 DN 2s | # | 0 | # | 10 | ws | 150 20 | 20 | 30 inch) 0957 1.50 2057 3.067 4026 SHS 6.065 7981 1002 1200 dinm) 2664 4089 5250 7793 1023 12 1941 27 245 3048 f 00225 00202 0.0190 GOI73 0.016 MSS a0149 ool 0834 0.0130 ‘multiple! 00482 0.1247 02441 05632 L000 1.809 23725 42250 61205 9.9509 Optra tor -30F 33 «R666 ‘Table 93 tom 307 «197341452 50D 0003480 MepailiOC fe : Opimun’ —«9D—« BS AGOS«NOGD SDSS SRGSCSTR tons + 4CF Table C93 tor ACF 43 171-328 92S 19032 SSB_CIBS MO. 3305, Ye psil106 fe ‘Note: Toe multgier is A/f, lative to 1PS4 (DN 100) «t 1.0000. ‘Note2: Optimum pact is from Reference &ca7o PIPING SYSTEMS selected, the equivalert lengths including the valves aad fittings are totale¢ from ‘Tables C9.10a, b, and >, The pressure drop or tempercture drop can be found by using the formulas in Notes 2 anc 3 under Tables C9.7, C9.8, and C9.9, There are fectors othe: than economics, thet affect line sizes. These factors are discusse the following paragraphs. Suction Lines Ir Fig. C9.1¢, the evaporator pressure between points Sand | is established by the heat load in the evaperator. The compressor pumps the refrigerant from point 2 to.point 3. A pressure loss in the suction line between point 1 nd 2 causes the ‘compressor :0 operate at a lower suction pressure. Sirce the vagor expands at a lower presstre and sirce the conpressor is essentially a fixed-volume device, @ reduction in suction pressure causes a decrease in mass fiow. The increased pressure differential due to pressure loss also results in more power per ton (per kW) of fefrigeration, Pressure drop and cests are not the only consideratiors in the selection of suction lines for refrigeration systems, Th: chlorinated hydrocerbons (CFCs and HCFCs) are miscible to some degree with mineral oil. The newo: fluorinated hydroca:bons (HECs) require synthetic lubricants in most cases to provide for mutual solubility that will keeo the oil in circulation to enable the retura of oil to the compressor. ‘The sizing ofoil return lines for vertical suction risers is dscussed later in connection wih Table (9.13, It is custemary to size suction lines so. that the total Joss.in pressure doesn't exceed the equivalent o” about 4°F (2.2°C) in saturated temperatures for halocerbon igerants ind about °F (1.1°C) for ammonia. Greate equivalent pressure drops are toleratedin very long piping eystems, Tables C9.22 ard C9.2b roveal the rclation between pressure change and temperature change for common revrigerants. Keep inmind that the system capacity less, or the increased cost of refrgeration die tO suction line rressure los is about 2.2 percent per°F (4 percent per®=). Tables C9.4, C&.5, and C96 show suction line capacities in tons for two tempecature loss rates per 100 fect (kW por K'm). The capacities are shown for single-stage systems with, condensing temperatures at 86°F (30°C) for R-717 anc 104°F (APC) for R134 22. When the liquid feed is precosled in economized or two-stage systeins, the system capadties increase in proportion to the enthalpy increase between Iquid and suction. Tables of the thermal properties of the refrigerant are required to assist in such an analysis Table C9.7 for ammonia lists suction line e¢pacities. Tables CS8 for R-134a and C39 for R-22 list suction line capacities. The formulas ia the table notes provide the means of adjusting the line capacity or temperature drop frem the table values tc actual values. The factors in Nete 4 provide for the effect, of different condensing temperatues. ‘The application of two-phase flow or the simultaneous liquid ard vapor flow in return lines irom evaporators to accumulators is not ‘reated in this discusion. Suifice it to say that the pressure drop in horizontal two-phase flow increases the pressure drop approximately in proportion to the incresse in the mass flow. This cans that the velocity in the suction line for liquid overfeed appications stould be reduced approximately by the reciprocal of the square root of the circulating nunber. The circulating number is the reciprocal of the refrigerart quality at the eviporator esit. For exanple, the velocity in return piping, for a ciraulating nunber of ¢ (vapor quality at the exit of the evaporator is 0.25 or 3 parts liquid and 1 partREFRIGERATION SYSTEMSPIFING ea71 TABLE C9.4 (IP) Suction Line Capacities:in Tons for Ammonia wih Pressure Drops of 0.25 and 0.30" per 100 ft Equivalent Saturated suction tenperature, "F Steel line size 40 A= 0057 a¢=050F | a+ 025F A= 0ser | ar 02S ar =050F wrs | sat} ap 0046 ap = 008 | ap= aor ay =0.35| sp=aiz3 ap = 0245 |» os os 006 ‘09 oat ous » | a 006 040 02 048 02 032 w| a os oz 028 oz 050 073, 1] a 030 oas os7 ot 099 14s m | 4 02 421 133 224 265 384 m | 4 Ls 133 wm 338 400 580 2 | aw 28 357 454 659 709 1126 mm | a 3M 578 223 1056 1250 wos 3 | # 71.0 3030 00 Figs} 223 3209 an] a 147. jazi} 2 2681. | 3862 4566 a1 3 fH} 266 ~ 3B. 4868 | 7007 220 11960 6 | « as ax as ® § 41426 13437 1.44 8 | « 9007 19.19 wads 73s3n | arin sors wo | « 16436 2639 | ISL = ann 50458 723.08 of | ase 30.88 ass 00610 | 0033 uss9 ‘Sturatel suction tenperature, "F Steel ine size 0 2» 0 [| a= 0OoF ar 0307 r= 0asF A= 030F Nps |'sca| ap =0184 Ap = 0368 4p = 0366 Ap = 0.582 % | 0x 026 oa 033 we | 0% 052 oa 105 % | 0 og 118 197 238 i | a 12 234 38 459 1% | 40 an 621 9.8 sae wm | 40 68 934 48 10864 2-| 40 ne a3 2as 3608 m | 40 2.9 2884 4537 s7st 3 | aw aa 3135 s 80.8) 10193 4] 40 Bo 10517} 13616037 eect 20834 3) 0} ime 19055 | 7353 29097 | 29688 37618 6 | 40 | a6 30862 n039 460.07 | 480.95 coos7 B | 40 | ass 6382 | 1699 9207 | oesss 125034 0 | 4 | s0647 suasm | 128696 maa | 79655220399 2 | | 20% 183928 | 195456 779037 | 286223361323, ‘Note: Capacities are in tors of refrigeration rsulting in a ine fcietion Iss (Ap in ps per 100 & ‘equivalent pipe lengi, vith corresponding cnange ,ar in F per \00 ft) in setuation femperture, vapor) should be reduced to 50 percent of the veloxity allowed for the vapor flow alone, Vertical rises in two-phase flow requir: higher velocitizs or multiple risers to carry the liquid upward with redueec pressure losses, Preistre losses in two- phase flow are redused when accumulato’s are located below aad close toevapora- fors. Wh2n accumu ators are located above evaporators, arrangements for separat- and pumping the liquid return improve efficiency,esi oor ¥5"180% orosse seuset ores IS St vouort aera ose enzont Dre e609 Boe od sr E89 ise weve wut ost evo wee orist sor set eee wor 0566 sss or ocetr ooo veer wie 08 we we ra Bst 9 wee ores voor 16 os are itt 198 ra or ser TER ue wre * 159 we ue st se tre 30 wi wo o oT wo 590 «0 sr 0 Iso sco a oro or urea zac = dy wpa vor = oy wea psc = dy wea vor = dy yea rye = dy vers dy Na upt9o= 10 wy 000 = 19 wrx 10 =19 wy soo = 19 e100 = 19 ‘U/y <000 = 99 Pano oe oF os- as au pag. ‘Dy santeieduey wopione PeTEENES woqeanbe wyyt 1070 pue co0'9 30 sdoac{ amsroig 103 tuoMULY 10) (A\3) SaNPedED aUrT UORINS (NS) WED 37VL care(wor 11) enuf gun rad oxmpexodiray dar ops donsrofyar jo sromoy urate sompEdig aN wonernyes wy — Surpeodson09 may ‘Gina dey sans yom 239 so a ‘s0ni0r Leo v 006 66819 1696 ose swseie ooeerz 00@ scout L990 oct welt oj scr | 1 set s6'T@e Lagse Poor cre sent 8 wu ecR6 9 wa01 se 6 ess ete o% W698 sore we wer oror st 56 1s a wp ose. st wz é eT or ua essi = ay wpe 9% = av wrates dy... wna ye = dy wea tar = dv Na wp 190 = 59 ws s000 = 17 wal enn = 19 wry v0 =e ms £000 = 19 muon oc seg 20, reais st caraines us stueys = 17 ‘np dap simeoid = ty ovr — owt | oss = ovort-] Oe =e «| oven vos | op | 9 seve oosor. | tow use over 0869 076E oss ook |S ost = ogss. | ogee. Or wee ore ose | ob | oro = os | oser over oer O68 01 ovst ov | ost E91 see Ocer ESL vor 66S es | Op | Hz ot, oscar owe ee we m9 au ses | ae oes of wy we we +61 ue | oo | wt sre +s wz we or 1ez ert wet | op | mt #60 +e wo oor 190 $80 ovo oo | og | 1 sro 690 60 950 Teo sv sco seo | on | tev ore 10 sco v0 zo Ito sro | | Hos | san oss ootez | oover ooust | ors == w'ser | con9 08°86 oozur vost | oes ores ores wee oer oc 19 ores ose | say oes 066 rey osez oer oror = ows, | ose ose ote = woe oor weve ore vee | uc ose ocr = woe ort wear oo! ovr} ower oes 3, wer 69 oor 65 oor! ect 6 vos, el we ws 19% w9 suet ore we ose eT see 16z se oF7 wz wT Re art eT oor 7 sel os 980 aaa 0 660 20 oz 990 a0 wo Bo eo ro co) sro sto «zo 910. Zo wo sro v0 *20 eco 9x0 800 zo. rian) ora. ~wo=dy wo- dv] wo=dv eo=dv| ceo=dy o=dvl ocv= ay o90=4y| so~%y oso-a| ao ASONIF dbs IV | bSO=IV al sv [aso= 17 star | bs0=1V ai=1'|aso=17 atasy | ‘ddoo ‘dk on of - @ or 0 7 a. ‘sunyeradiay uonons paremeg oz oury Guoneoyday aBers-ySiq 20 -f8uis) epg] wopesafisjay 303 suo, w senpedeD aur] wonDNS (dl) $760 STEVE caraose noes {Teor onst oor oot oust cee 059% co oes sas ceo oe ey 9 wee arse seaee vet os ove 9682 ort oer o® wet veer ws ice z 959 sE6 BP se se we ‘er we sez e sr see set ET st 0 wt wo 90 or : . : Na - : : feos wo SLE ost wee | arent waz cozt cost sor wu wet | ore ot | cos oro ‘6s wz 6 cee wet | oss ores ‘Ore ws9 se ores » wr ae o0e ore west a (soe coe ts ore orze ost vee wa oat ocor st o cet wet 6 orer we on a9 6 se ee wo ws tet wy o9 see wr % BE tes cod ot we se or we az wz we et az ar et 60 wr at ot wet 0 at Bo wo 50 wo st Bo 160 #0 wo 50 wo 50 wo zt a “a9 nado, ger= dv sine. | isa ocedy | ota cae | weedy cede | cney gisdy | os Aworw Awvmw |xwo-w xwo=w|xwoew xwo=i | xico=W xeco=" | xWO= XOOA~1V| yermdy oO sa o me fe or (, ‘simezoduay vonoed Paes | : Ce woparabe wy 190 Pe 20 J0 sdorg aunssorg 105 (suopeayddy ofeag-y2tpy 30 -o78uig) wECT HOMBIOBIIA 105, SMEMONEN wy safiiaedag SHY] Non=NG {19) INS'Co STEVE e478,2 €OL #24 4, “doxp aunsseid oy Supuodsasioo aumesoduias arenes tn 1Ey> = 29 ‘wilde, omy yeojentnke 1 oOT sed 982 “wor aby, o% anp dip amend — ay sseot ccisct | owsee ze mes | oi Cats «| wie ~~~ | a) sis toeeet | tezre cui suse | sorse tor | Succ; ase, | ae | ot sive Tae | cree woe roves | est. = arize | Sose bree, «| or | ous taste, | cot west ieee stax praot | ous ws | oF | 8 seo aso, | aut wu srr | sor exon, | ree «=e | | res | tees or us| + tom «sve | cor Sees | | reac Ser See | teat me | toe an | & | & eit ost esor ose | at wo | ay of om | a rt bv toot 39 ws is ire wee co | o | 2 feu ero ats se fos ce ae sr we | oo | se ser 69 sre oe ist ser ez 560 rn oe rer rer oe er oso 4 eo wo | of S60 st 0 oro = oe oo ar i | aw | % aro oo eo ozo «) ft sto ar zo | we | eo oD ty or S10 too ove 00 so | | wos | san pais wie wre sw uo | sew uw | seu aese i ws 36 wie oe | om tig | sect, we See Gee sexe | ret sor | ae cot we wie Re wa us| ee wa | as oe we wst twee set wn | ter a3 see ier we ie anit ee ors set ie ar aot a sy oe a set at wz 580 wa oa nt woe or i os iso ar ri wo vt fei i oo ao 160 ov a eo sco % sro it es ie wo wo i io zo * iso ko seo Ba ree sv = ao Ho % we ore er ao are 0 a so i a iero= dv seri ay | es0= dy sro=4y aso=dy | @co=dv uso=dy | wto-dy tecv- a | a0 aso=w gai | asonoy sony re |asomw Tansy | iso=iy “ata | to iL ow ° = o- : 10'0 pur z90 70 sdoig amssarg soy (sxonvonddy ofeis-uBti 30 -i8ure) 77 wexoBiuiex 103 SEMOME UI Saino aMrT HOMME (1G) HOO FIV a7?‘souran%9 £,0P+ 0) Op— soy paxdde uayin ye wey ss poroyye s Cypedes ag, gy alntesadey soyeiodeng wonons 39¢ Wo posed ae Sahipacea aut nb] paw SURES Aron 1 a we or ort on % su wopeunbo pus rade eo ol oe 7 say op Pr m a0 501 a (gas ume mt ur smerediaoy mvvReey Tal. s8ieU9HICt uonots “ausuepuo3 produ) uo ) ete p= Aden zamyes 9410 20) Ayeodeo aur “z Junpiedun Bannan mnquonay uouy UNO 001 24 ay jones u a20eu3 Buspuodrs0> = 79 aimsoduo, Basopuoo sowie ic) uoIse 2UMNO|O 9-64 SHOE OHI ‘toy aun sqeunbs fo ‘oqeuy eo; anp dosp sunsaad = dy ‘dm “arvessdu, Bostepros 3.96 to posoq axe oye ah C1 sone e1984)01}0 stor Wa syed Sg “1 oN = = a | a = srs See reise Swett =| «Yt - - o | o = ORC Shue HEOT © Tuer ov | or = = or | 8 voser 9991 106 995 tue | ov | — = or | 9 os su ceey 91% ovst o | 9 > ~ or | s 60m vz hue rout o } s 12908 ere | or | + ose eae VOT ad ov fy suse sor | or fe oro our se 19 ow fe preset wre | oy | me re cep sw ost o | 9 vrs ve | op | z Bus o6E os zor z cove sur | oo | at ow sor va 8 % owe 18 os | mt 661 oer 3 9s m oor eo oo | 1 ot cs ve Te = zr 97 we | % se oz = = = ove eer os | = = - ~ - ra 98 w | % = = = - ~ acos ay “wasoot | wos | sai wi=dy st=dy wo=dy sro=dy tep=dy sdoz=av = Speen Fo | kl =i or o 0 w~ or— soa prabyy als ai 2umezediuay uonens payeimeg (9 oot/at = 79) sous woreng (soonearddy 08s witty 20 -ofuys) ejourury 203 suey, ut SoUDEUK DUTT PHOT pur HIMBA “uoNaNS (al) £59 STAY care( eae) “Tatas Tnedeo romipt,) Fr ramay/V "981 = 7 °7 tBu qusyeaqnba pov tyaedva Jomo 20} 19 aImLBodweT WORRIMES “¢ > pe aBIEDEIC S ‘wonans 9,5~ U0 pasey axe sonpedeo oxy ui sai Bp odtasa05 uaa ‘wis dtd woqea pines tosessdajaro srewopy uo sapped ‘eompnaeduiey2ujopwos seq ao) avn) Ba ‘dayryy “sinesadan Fasaspus evo _ orn SN ‘coes: Tue ——«GWBSST ——svelae—=~=CS clas «—=«COHOES~C esus yom —Eeklgs GEES SUT eserar worse, “Semis ou + cow. eescte suse .vsoe rie ose a0 rue “deer ms greet a es got) eet eo sac bao te 138 2 gate: set ree zs vse % eat, * 908, ese re zor o we ee zoe 961 rat ze +9 ose ont ¥6 ss © ea cet ve or oc o m 59) oe x4 wT st 8 i 6 a 0 a WR O5b = dT ‘ sac do ovit'-dy gwi=dy eott—dy 6 = dy woucioe=17 stot fs o OO anon) weno POO wed ovs9= Fy w/Z00= 1 WI 200 = IV ou oRaNeG, ‘501 totons (suopmonddy ofime-nfier 1 -offnyg) wmonwy 203 ssnpedey 2urT pINETT PUY “BsmyoIg ‘woRENS (1S) ECD TTEVL caracose = cars! oor v9. - > - cons onses. oozre: ~ vos wun were OoleE ~ aaeTT os. us cores onset uP ooo = ontst ret OTN we ovgs9 onset + owe coz case ‘oe o8ss Yr ors we zp 000 we oss ozs ores cue =~ aoe u ee we oo 06e 0669 ur. nese rue ose ose ower net we OOEST oesy we one onze cet octt = oor tee ve ove ors wt over oxst us 69 ors, to ver 00s ‘ore MT oor 66 a8 wr we. ase va over ora ur sos as we wz Ri sv va oor won % ez a scl mt 160 10 % ws. we % att +990 oro feo uzo w% ez, ez He oo eso “seo eo ko 4 wendy wdicor go ov @ 0, [veer Whe ert ont £0, min = PA spdaeD . s3idos TM i ‘sumodura} 7 = oatrisdlg 2upuodsaix09 Taiy, pure sero ouons pareumes: c {ait *s 228 OUT ~o of e or o eas our] $00 pmb] (a onrnsé ez = dp ‘a.t = 19) manele! coca Simian enero pareinieg (5 ort = 1) sat womans. (suonesidey sieisiiy 10 21s) rT pus “eBuequsny My (et) BU STE Obcr eesoFusoUr 205 suOy, uy cemordes our e480“ou soo) tin hoon yum sucrvonnny ‘320 TeNUESens 8! Butfooogas 1960 ao = 208 TMEV) cede op = pesto on steppe i aman ar 19 roaeaduay wones 0 29) Kipede> ourT -Z ’ ‘Woot ody, “meso vonescis x arms dmpeodssio9 = 17 “simyesohnny topaopucs spo soy arora Baaayoy ow ka Duisedeo ager ‘oot aut worn jog ooLsod cogs Seer onpdospameand = 2 ‘aire srvenes Bocas yt ae Foray oan one SOE ‘uoyesa9.g 70 S001 aye ep aR, \ ON oes coust | o |» | ove -onoo oven | opto ose os ocr] | ooste coset | or |) © |ooszs ose onset | oper wth 065 or » | ¢ woge ‘OvOL OF we ‘avo ‘ose ower ove ores ocst ov we cost —czer |» fz | vce — ooae oye’ | aver. ovr 356 | ove 6st 8 A ‘orst Jose aro wl sor oO “LT ogty ogst 8 *L oror 996 a ‘os ore oF a” oot osot. Joon || |i oe we Be wt set wm | 7t 165 169 om tet at srt aso 590 mw | % ace see ox | % | seo two 0 eo co om | . Hos | SdN HOS ‘Sdn PS ris castcouse way) wm ws oH. 1 “dos | oo oveie were sor onOLIT O09E 006 ose Oo'ize oes oorlz ower ooLrT 6h sqroant oraz | ose’ ooute — cowwe- | 'covat cow ovstt wre 2 009 over | coer weer out | ees cru. orvg wes ts ever ace | orm ons’ cous, | San woe ovee’ ose 2 cost oso | ose ore apse | ores | ose oser orsr se 565 cow | orm oroe” cet | oer D,oh UO paseg a1 [EE OUT UT SINIEA “e " cuammraliyat jo eneunary oy a0 oere oozes, ovees | ocese ooze oes onzee vor coors ov90e ootse | ovez coast corer on601 08 ows rect ovest | oooet A oees ove orto 9 on 3¥s oc s0t oroor | oree. Songs Lg opse ws ones ors oss ‘oot ore oct ‘ono ow cosa osze SOE ore wor orst ore © on't6 oat O9T ocr vy WE ee sv Fa ra 16. ovo pee. 2 we ue a we ove se a yee 9ee 961 or sw oezt we we oot a ‘or a0 om . SaN Betscept 1261 vse, rise | ever nL cd LS68 Suh ree rect | eLer v9 ety 09 S10 601 zoe zoct | 6% Om Fee SUI od 53 ss vse | ess sz su 6a oeiz O68 ser sey | cee ost oor =z sat se vee oe | 081 slo Bet i) zu br ter | ror tye 6 oue ze s8 se | gs woot ot zl ge zy ce | 6% el 90° 150 “9 te ST v1 tL so 0 Ei e $80 sco | 90 +) oS se=4y uid oor ov Or— ier or srt 640 dt=W = PA | Tae ay ‘oamqosedusoy a oorrsd ‘2 Suppucdsezi05 ee worons paren 9 Pa § Seno 22S want Ta = mrs > 0 « Ce sour] pmnbry . ($9 SOE = dV ST =79) compendia, ws oe ts ae ‘2imjeseduie, uoons payeameg (3 00t/.2 = 79) Soult uonongs tL (suonvoyddy a¥e. er weRBuyPY 205 svoy, wi sanpedeD ourT pinbrT pue ‘aBrey>eIc ‘woHONS c.484“aro san “sun sos eatabor Cour sour FuoT {10a s0 Buy 2283 2q 4eur aunyetedu gh ion ga suoyronddy ‘pon oa Aru ou of joneus w uow st eux “ao tesvooueo Buna! aio pnueiges | BurDoogNs feaneazaeuon s dy doyp sunseid om ou, «asta J2na900 9) poteiousd Se lus ten popacasons H UMmoYs BUEN U5 eroduoy uoneames “¢ (HRY atone. = 4 x (EY aah = Ape 20 ‘aur wong. -somesodoy Bassropuoo so scp svo;ey Boos 3p kas {Aidainu ‘arnseiadinay Susuapues w,cor uo pameg 2am 219M StU SARA, cron cd ov v som | vet ott sats tee | €0KS: sar | or | “ROT | TOS 89S OE ee zOL |b ss0e 9 ov | vi srs || ‘pS ee ree’ eto |r oz6r wes ow | z Voor || ese te wet LS |b 528 ee 08 0 ert ‘sor Lee | Ts 90 os | mi gu |) 66) OL By Ee Ot |e ra en oe | 1 ro | ee we ez" et ceo | oF sz 69 os | % ‘eg | .ge8 gt et go oso | ob cs se o_ | s a seo" 00 |b Hos | sat HOS Pais paiso69r ovwor el | eth = gues zee coe ase sor oot csr lr | GLEE Tuyen o Osthi ore vor | e812 sest 806 = we TBU 2 ere str cust | cer ope ives Sez 6 E666, 00706; wel | eID Lrsr © soe ORL bE ow sone? Lore wor | teu cee wast ovat cs se ole ocee woe | peo eH RS LE x 8199 wst wer | reor tse eh ee z ore 68 ew. | sss ott OWT oT sr we sow wey | eee sve wT S60 190 st vel owe we | set se sco oso ze a aun adda wreaor=dy sco st oe op Tek us ale ue, 960 Sermon iy = Seen | 5, “dusy vonons payemeg yea ‘dy Bunpuodsaizoy, ‘souy pinbry 5 (osve = dy “wrx wo = 79) st se no OF usta 2. ‘simrersduray uonons porsinies (t/a #00 = 29) sou uonong (cuorwondaly afexg-qai 30 -y8uig) 77 tweso8t1a% 20F SARMOTN UF SepPARAD SANT PINE pee OFrOqaCIL woNaNG (19) ING'OD STAVLtout Se01 fan's Seipeongng M908 TCS W IONS 5p 9 100 6 dy, doup oxnesoud 30 2414 r0¥9H Sumtonoy 2en hq sapped 21Gb ‘a0 UO porEg aye SIGE aM) UH SOMA ‘soammezsduey Supsaepuoo 19 Aqdommy simyeiodun Susu ct9. ELE vu 6zer 9689 voor ecw OBIT os bo oan 1901S ‘sanmroduay Susuopuoo juesofuyar om teu senso Bt Bsa HOSUaPUOD PTOI TOI po s0f 17 amesoduray wontamis “¢ ) frcede ag = sete ery 17 sxmesaduiy uowemes 12q0 30) Aypedes aut“case TIPING SYSTEMS. Discharge Lines In Figure Cs.1a, the condenser pressure between points 4 and 5 is established by the temperature of the cooling medium. A pressure loss in the discharge line, shown, between poiats 3 and 4, increases the pressure ratio in the compressor, stown between polats 2 and 3. With higher discharge pressures, the compressor must pump againg a higher pressure differential. This increases the power required, Teduces the volumetric efficiency, and reduces the capadty of the system. Simlarly aswith sucticn lines, the sizing of oil return lines for verical risers vith halocarbon rerigerants & discussed later in connection with Table C9.14 for discharge (hot 8) risers. Itis desirable to not exceed 2.°F (1.1°C) equivalent loss in the discharge line. Greater equivalent pressure drops are tolerated in larje systems The disctarge lines are listed for capacities at 1° per 100 feet (0.02 E/m) in Tasles C9.7, and C9.9a. The informetion in the notes is useful in adjusting the line capa and pressure drops to other conditions. Liquid Lines ‘When a system is equipped with arefrigerant receiver, the liquid line entering the receiver is ually sizec generousy to assure free flow from the condenser. An allowable velocity of 10¢ ft/min (0.5 m/s) is typical. However, Table 9.15 provides a rational sizing of condenser drain lines when the line also provices equalizction beween the receiver and condenser. Velocities of 150 percent of Table C9.1! are used when a ‘eparate equalizing lire, from the top of the receiver to the condenscr inlet, allows the condenser drain tine to run full of liquid, Piping for condenser: is a comrlex subject, requiring consideration of presure drop, trapping of individual circuits, subcooling of liquic, provisior for purgirg of air and acconmodation of temperature differences betveen the receiver anc the condensers, among other issues. A single condenser coil can be ecnnected tc the receiver without an equalizer, and with no liquid head H provided the drain line is sized in accord with Table C9.15 capacities so as to ‘liow vapor flow from the receiver back to the condenser. Multiple condensers can be connected in series, but are usually connected as shewn in Fig. C9.4, This parallel arrangement requires that the comienser coi be of 2qual size, that the waits operate together, and that the receiver be located in an ambient temperature that is equal to or lower than tke inlet air :emperature to the condenser, Some liquid subcooling occurs, as the liquid tends tohang up in the lover tubes. The amoun: of liquid head Hf under these restraints is only that waich wil induce flow. If the coil sizes are unequal or do not always operate together, ofthe receiver is not Iccated such that itis below the ambiemt temperature 01 the inlet air to thecondenser, additionalheight H isneeded tomake up fcr differences in coi pressure drops at fill load, or to overcome the p:essure equivalent of the temperature cifference between the receiver and the condenser. ‘Two methods of piping parallel evaporative condensers are illustrated in Figs. C95 and C95. Both methods require equalizer lines ‘rom the receiver to the cordenser inlet piping. When adequate equalizer lines ar» used, the drain linescan be sized for 150 percent of the flow suggested in Table C9.15. One advantag> of the surge-type hookup i. Fig. C9.5 s that subcooling of Iquid acconplished in the condenser is celivered directly to the liquid line feeding the evapontors. This can be ameasuratle increasein efficiency, since the heating of liquid in awarm receverREFRIGERATION SYSTEM! PIPING C483 ain=cooLeo CONDENSERS SINGLE ciRcurT Purse. yr wives AX Lecare stor VALVES AS ‘close 70! horizontal HEADER As Possiete E:LEG n SHOULD BE MAXIMUM POSSIBLE. SEE Text FoR MINIMUaAND” ** GMITATIONS, is bypzssed. Ancher feature of the surge-typé amangement is that the individual dcop legs are isoted from each other by the jiquid trapped at the liqiid level in the receiver. Alsc, as the liquid rises inthe drain kg to accomodate, inc-eased load and reduced presure at the coil outlet, the receiver can usually provide plenty of liquid to fill thé individual drop legs. When coils-are not of equal size or when some condensers may be shut off, the static‘pressure of the liquid height in the drains of the working éoils equalizes the pressure drop acrcss all of the coils and provides individual pumping force to drain cach condenser coil. With through-type receivers, shown in Fig. C9.6, the horzontel portion of the drain line should be limitec t0 the flow in Table C9.15, as the liquid volume in the horizontalline should provide for filling the individual drop kgs of the working ccils. The licuid head h in either case needs to be equal to the differences in pressure drops in ‘he parallel ‘coils uader all conditions of load. Typical height & with ammonia is 4 to 6 ft (120 ‘to 180) mm). Typical height h with haocarbons is 8 ft (2400 mm). ‘The design constraint on liquid line sizes between points 5 and 6 in Fig. C9.1a when the liquid isnear to saturation or et the bubbc poiiat, is that too much pressure drop vill result ia liquid flashing to vapor. The sbject is to deliver liquid to, the expansion valve, Excessive pressure drcp in high-pressure liquid lines caases forma: tion of vapor butbles or flash gas. Thit flash gas can-severely erode the valve seal cor neeile in an expansioa valve. Also, since vapor occupies a much grecter volume than liquid, the remaining liquid will xe forced along the’pipe at a much higher velocity. The increased velocity causes a greater pressure drop, which aso restricts ‘the mass flow. A rise in elevation of the line, o: the pressure drop due to pipe friction, may lower the pressure below the bubbie point. In some cases, a means of cooling the liquid may be necessary t> avoid forning flash zas. The liquid cooling can be obtained in an extra circuit ir the condenser, or may be provided withC490 PIPING SYSTEMS LIQUID ORAIN Lines “EQUALIZING LINE. FIGURE (9 _Hping ‘or panliel condensers wih surge-type receiver. suction/liquil heat exchangers ot separately Tefrigerated liquid subcooling heat exchangers. In condenser arrangenents similar to Fig. C9.4, there is some iquid subcooling a2complished in the bottom of the condensers. However, in equilized condenser arrangements as shown n Fig, C95, thers is little or no liquid subcooling. For receivers located above evaporators, there is litte concern about liquid flashing because the liquid gains in static pressure as it flows cownward, which provides efective sub:ooling. Fee evaporators located above andremote from liquid receiv- en, the pressure loss in the liguid line must be kept to a very Jow amount. Altrna- tively, the licuid may be delivered with pumps as is common in many large riping systems. In “ables C9.;,.C9.8, and C9, the listed captcities of large: liquid lines suffers some overstatement compared to recommended capacities, while the istedREFRIGERATION SYSTEMSPIPING cast cou, DscneSson : EQUALIZING LINE) FIGURE 9.6 Pping for parullol condenses with top inlet receiver capacities of smaller lines are understated. As a guide for selecting the size of high- pressure liquid lines, the velocity pressure in one velocity tead for tke natural reftigerents, R-717, R-290, and R-744 should be about 0.055 psi (375 pa). The velocity pressure ir one velocity head fo: high-pressure liquid halocarbos refriger- anis should be abcut 0.11 psi (750 pa). See the defnition of relocity pressure per velocity head in th: second paragraph under “Optimum Sizirg.””casz PIPING SYSTEMS Pressure Loss in Vaives and Fittings Two methods are used for dete:mining the pressure losses in valves and futings. The most convenient one to use is the equivalent-length method, which expresses ihe flow resistance of valves and fittings in terms of lergth of the iame size cf pipe, Tables C9.:0a, b, and ¢ show ecuivalent lengths for salves and fittings. The total of the actual run of pipe plus the equivalent lengths for the valves and fittings gives the length 1sed in the calculation of pressure loss. The flow resistance of pipe and fittings can also be expressed in velocity heads ‘+, where ote velocity head is k = (L/D). This mettod is mor: complex, but it nay be more accurate, In order to use this method, tie length of pipe that has a msistance af one velocity head is determined so that tie actual run of pipe can be expressed 6 a numberof veloci'y heads (X). In order to accura‘ely caleulete the number of velocity heads in a length of pipe, the fristion factor must either be calculated cr determined from a Moody chart. To accurately estinate the number TABLE C9.102 (IP) Valve Losses ir Equivalent Feet of Pip: OO Nominal pipe cr tube Cd Swing Lift Sein Globe Wye Wye Angle” — Gate’ check® check. % a 8 6 6 06 3 Globe % 48 9 7 7 07 6 ‘and % 2 u 9 8 09 8° vettical 1 29 1s R 2 10 10 fi ™% 28 20 5 15 15 14 nme 1% 6 MB 18 18 16 2 SM me a ayy 6 3529 29 28 25 3 84 B35 35 32 30 ery 100 530 aL at 40 35 a 120 sa “a 45 40 5 140 na 58 58 69 50 6 170 8 70 0 60 8 20° Sas 85 80 Angle 10 280 M5 105105 100 iit 2 32000 165130130 20 ‘ame 14 3008S SS ss 35 as 16 410 210 180 180 50 ‘gle 18 40 240200200 6 20 300 SBS tas 200 24 a0 320 S265 140 ‘Note: Lossesare for valvesin fully openpostion and with screwed, welded, Ranged, or Na * These lose: do nck apply to valves wilh needlepoint seats. her * Regular end short pattert plug cock vaves, wien fly oper, have same loss as gate valve. For valve loses of short pattern plug cocks above 6 i, check with manufacturer ‘Losses alsowpoly to the inline, balltype check valve “For ¥ patten globe lift check valve wih seat approximately equal 0 the nomina pipe diametr, use valies of 60° wye valve for los.REFRIGERATION SYSTEMS PIPING cass TABLE C9.10aM(SI)" Valve Losses in Ejuivalent Meters of Pip: | For, values of 60 wy valve for los, of veloxity heads in a length of piping, the Reynolis numberand the pipe or tube Foughness must be determined. Tables 09.112 and b show representative velocity heads er k factors for valves and fitting: ‘ It should be nected that sizes of copper tubes are referred to by nominal inside diameter in TablesC9.10 and C9.11. In al other capacity tables the nominal outside ° diameter is used. This practice is probably carried over from the industry usage of referring to ASTM-280 or ACR tubing by nominal outside diameters, ind refer- ing to other copper tubing used in plumbing by nominal insile diameters. Example C9.1 Example of Pipe-Sizing Process. The blowing example details methods for de- termining flow rates, volume of flow, line sizing, and pressure-drop for R-22. ny ig Bana eo00s ‘oor ost | ove = oe = oor 009 % ove oee ogor | ots = 09 = ore os 0 oor “ot ose oon = wee = wee ow aL ose ‘09% o8, 29 = 02 =) 08 oe oT ove oct ow vss a oer 2 vee ove vr ox ost 60 os = o9L = ost ove a 08 09 08 ve = ter = aoe or ost ost oor wee = vor oS ost ove 8 ov 001 ove. ose vy. 6 ose oot oor 9 oz 78 ose. or oT 9 vi 78 oer s 06 19 or out se ts out “9 oor + 08 65 ost ost cl ty ost 6s 06 we Ot os ost oz oO oF oz oe ez € 9s vr oz vor zs ze vor Tr 09 vie Ly fe oor ze sr oz re ee os z ve 3% 08 9 ¥e tz 9 9% oF % te ez oL 9s oe it oc 7 re va zz a os: tee ate cer sty “1 oz U a a oF Te oT %o te ca oz % vt or ot St eT $0 st oT oT ” +i i 60... | ee % Tr Lo & 60 vr % : : : uw mW Ww tH ay S ro) B® ‘one 04m % % one ww add poonpou pope Sommtear | [TS f : rourwon! a Oe aang bs sens. snipes as hoor [7 st 06 | -Fuoten6 606 noy nomadieng | -aola ‘vod ping quoouis Fi pBq WOON 3 5 : (suonasuuc) pezesg pur ‘pareja ‘poduely ‘popla, ‘peasssg) odiq Jo y99q woyeambg w} Sasso] SUN (el) FOL'6D TTAVL ca94“gt 01 nbs Krnramsoudde aay st a SE 6c _ v6 _ a st €1 or OE St _ 6L _ OL 7 00s a x8 % w -— - ss et Ose Fe a @ |e) @€ = 2B 5 & oT] te Bo POR |) Boor LB. > OB | Oe wo @ [oe |g = we =. & Bl ee ey OE 6 OL “OF ve! ot OE or Ost oi ee Te Le lee Ot # 8 |p |e 2 om BF of | 8 eet @ 8 |e |e 8 RB A 8 of |e oe +0 £0 60 80 +0 70 80 £0 so SL orl la LU UU | wwe Oe |S G G8 QH Bl ee eo om SS ma retin pen | US ae N x PS gang PSPS ps ws oma O81 st sh 06 “8407 06 “06 ca aerate a cogs NTS 5293 pag qioows (GuoqaauuOD paZtag puD ‘PereLY ‘POBUR “PyPIHN HomWiWy) wig Jo seIEHY WaTEoMbg Us oemoT Sem KIS] AVTOL'CO TIONS 495arouNeIp ysayfeus Ye S90] 10} OAR DIL BION cect oer oes ow a % oot oz oo oat = 24 Set o 006 ost oes ost | ow = foe = 1 Ol 096 0s 05 ost” _ _ OST 7 _ or 9 00 os os ot, = = | os > - #1 ous oe. ove oe om ose oer ow = a O9r 009 O6e ry Otr s o0z _ om Ooze _ or ose our ove oe ee es 8 oe occ oar oe oon ust | see 5 00z Ow OvT Ole os 06 oer os ost Ow $s oor ove oot ou, | se 9 oe | ge) Ono ‘ oer OLE ss ou Of 09 dt oc c6 ost we on on zz oF s.r sf ee £ ts ot os ou owes | oe) se 59 06 vo gr! oe oe | ae z os 99 ef 99 er te 6% tL oF as HL ee es St eS oot ee | ot oe OS al “et ut st “ee £0 a1 oT £0 oz ve 1 tt 8e vt st $0, oT cL so st st Me st 8T oT eT vo £0 60 ro UT st a” TT st 80 st go" so a) £0 30 vl % =? P= + = a= ‘sas 2qm 3 =H a wo odd ra) es em al ett, wey souenug, we ‘souBaUEY % ” % % % ” ‘wonoaford adig ‘apa dieys G/p “wors2nu0s ueppag fp “yaouraieju2 wappng dig Jo yoo wopeanba ur s0s507] Sum3 [ees (4) 20160 STEVL cass-p sopaunap joes 18 sess 10} 2196 JO} ON iz4 6 st ra ss _ _ ss _ _ or oF a 9% a3 60 st x4 60 gt oF 06 so. so £0 so To £0 0 vo €0 so st 3 sal|dal Ss 4) = | ‘aig Jo Sia WoyeATTb WSO eds (Is) W201'6O T1avL 4976161 1H) 100g HOC BeyoHndy ae:n0s oe) OY ro os re wo wee «OLD + Tt pro! 9 TE 060 T= weOCED og” € wz STO 9 €f 0 SBT oO seo seo et ot rok vr 60 OT e0. sO OF z st gt tO ot oer ee. zt wat tt 9 OB Lt 60) eT feo oD “a of 9 HO | 6 st 60 ST vO LD st 1 ee 19 OO wT SEO HLT % ss = 0 ew OT % os owe eS % aajts aA aNBA age WLI aU page — oz i 9rd 9 0 m0 co 0 OS CTO > oz zz wo Z 900 so co 89 sro. szO_S HO é oz eo a 60 iz0 sco ero. OS ue oz st we 6 180 00 so = DOO z oz oe = or (O60 s€0 wo =D SEO FO al oz ee = a $60 $0 wo zo co %t ow a = a ot 0 wo wo weer c ‘aniea NBA Avs. ‘Ava qpueg ‘Fuo} Bor ‘Suo1 ‘Buo} ‘Ban (San) ons pap aiuy DAD. meq mg tT aiid auyms wumey wn Sy 06D IBURION sSumng odig popeas peSuel—stoes Y, ILL TIAVL casa~ 6161 TED yoog Ona Suyzausiiug :2omag or so oes hen e070 of oor ot fo 0 1 et wo 9 060 oO 10 FOB or so 0 te ro" §9 060 oo seo e098 o so 90 eT sk 060 io. wO Or os o so 90 «st gz 8 0650 zo. 0 tI. ob ot f0 sy et. 8 060 0, $90 cl ee ot so 990 oe =o 6 060 veo, 80 St ot so 90 veo 060 seo, to tte or so 90 ss 060 wo, > et on so 90 ok 060 Ce ed vera eur nea aaa ouy pus Boo] rN aa papafory sens qmow poy offuly 0L uming Sy Eid Pa Fag "06 0 © preruio © ‘sfumur sdtg pamarog-—sioned.¥, (18) (WHEL H69 TIL.“GL6L TH) 390g HMA Bupiomsig meg oCe~«OS*«S 0 0. #0 ~eo 9ro.~—sero.~=CweS*~« 4 Te 900 “s. 50 600 +00 sto 900 +0 $70 0st oem 0 oro tO. tee oe, re po wo reroute oe wo 80 stos ost 4 tt wo L 0 0 sto a 800 so v0 OR 4 €t it0 3 0 8r0 0 sco 6h 0 seo 9 o 0st wo 0c——siDS=“(tsEDSCSsDSCwDSwDS ot of 7 ot 60 £0 sO OF0 10 se0 oro OF oe a = 0. sto ick oe tt oreo ce oye ayes —aaea Cea’ Coe SSCSR)SSCS SSCS SSCS yom auy, DBO, pega tid Suyes UN wy St (06 006 [euywoN, sdumty oda pepiag, pedueyy—sioweg ¥ (1s) (WITL1'60 TT8VL 501c.502 PIPING SYSTEMS Given: A refrigeration system uses refrigerant R-22 with a capacity of 100 tors (352 sW) at an evaporating temperatare of 23°F (-5°C) and a water-cooled condensing cemperature of 104° (40°C), The suction line is 82 ft (25 m) long, with two angle valves, one swing chek valve, and three elbows. The discharge Ine is 98 ft (30 m) ‘ong, with one globe valve, one lift check valve, and three elbovs. The liquid line '$ 180 ft (55 m) long, with two angle valves and three elbows. From Table C9.1, the refrigerant flow rate is 2.96 lb/min per ton (6.36 X 10°? kp/s per kW) Flow fo 100 tons 100 x 2.96 = 296 Ibm/min (352 x 6.36. 10°? = 2.2! kg/s) Suction specific voime at 23°F (—5°C) from table of thermoplysical properties: V, = 0.4856 Ibs/ft? (14.19 kg/m Actual suction volume: 296/08856 = 334 cfm (2.24/14.19 = C158 m/s) Approximate discharge volume from Eq. C9.3: Absolute pressure of R22 @ +104°F from Table C9.2a (IP) = 222.5 psa Absolute pressure of R-22 @23°F from'Table C9a (IP) = €1.21 psia Approximate discharge vohime = suction volume x P,/P, x 1.2 262 X ((1.21/222.5) x 12 = 36.5 cfm (Absolute pressure of R-22 @ +40°C from Table C9.2a SI = 1534 kPa) (Absohie pressure of R-22 @ ~5°C from Table Cé.2a SI = 22 kPa) (Approximate discharge volune m’/kW = suction volume X P,/P, X 1.2) (0.124 x 422/1545 x 1.2 = 0.9406 m’/s) Liquid density at 194°F (440°C) from table of thernopliysica. properties: 70.44 Tift (1.128 kg/m’) Actual liquid flow volume — mass flow/liquid densty 296/70.44 = 4.2 cfm (2,24/1,128 = 1.98 m/s) ‘Suction Line For the suction lise, try the nominal N?$ 4 steel pipe: (See Tables ©9.10a and b; note that the IP and SI valve and fitting Josses are not exactly equivalent because of rounding.) : Length of rur 82i (25m) Two angle vaves = = 94 28m) One swing check = 40ft 12m) . Taree elbows = 201 | 6m) ‘Total equivalent length = 2361 ft “W71m) Suction line temperature and pressure drop using Note 2 under Table O29 Interpolate tons capacity at +23°F......~ 123 tom (—5°C at 433 KW) At = 2 x 26.1/100 x (100/128)!5° BSR (LMC) Interpolate pressure drop at + 23°F ~ 2.32 pai (16 kPa) Ap = 2.32 « 236.1/100 x (100/123) = 4.33 psi (30.68 kPa) ‘The NPS4 steel pipe is satisfactory for the pressure drop. The rated pressures from Table C9.20 for ERW pipe is 732 x 085 = 622 osig (4277 kPa), whik theREFRIGERATION SYSTEMS PIPING c.503 suggested test pressure for R22 in Table C7.19 for the low side is 187 psig (1258 kPa). Zither seanless or ERW NPS 4 steel pipe is satisfactory for the required ‘capaci and pressure. DischargeLine, For ihe discharge line, try the 3 IPS steel pize. (See Tables C9.10e and }, note that the IP and ST equivalent Jengtls are not 2xactly equal because of rouading) Length of line = 98it (30m) One glove valve + = Gv (Om) One lift check = it (21m) Three long radius. ebows = 15% (4.6 n) Total equivalent length =251't (766) Using Table 09.9: The capacity of a NPS 3 pipe is 116.4 tons @ 1'F tempera ture drop. ‘At = 1,0(251/100)(100/116.4)1° =186F —1.03°C) Ap = 3.05 x (244,1/100)(100/116.4)!5 =56 psig (39.2 kPa. ‘The 3 IPS steel pipe is satisfactory for pressure drop. Forstrength, Table C9.26 shows the pressure rating as 664 psig (-564 kPa) for ERW schedule 40 sipe. Table C9.19 shows a suggested minimum ficld test pressire of 235 sig, (1628kPa) while Table 09.18 shoves the ANSI minimun design prossuro as 208 poig (1433 kPa) Actual U.S. industry practice is to use. 250 psig (1724 kPa) design pressure for water-cooled R-22 Systems, and 300 psig (2068 kPa) or higher for air-cooled R22 systems. . Liquid Lines. For the condenser drain, use 10C fpm (0.5 m/s) for sizing, when vapor from the receiver may flow in opsosition win the freedraining liquid. Table 9.9 irdicates the NPS 3 (DN 80) has a capacity of 118.5 tons at 100 ipm (412.7 kPa @0.5 m/s). When an equalizer line is used to vent the receiver to the top of: the condenser, a velocity of 150 fpm (75 m/s) may be used for full liquid flow in the drzin. A NPS 244 (DN 65) has a cipacity of 115 tons at 150 fpm (401 kPa at 0.75 m's), Forthe liquid line to the evaporator, try aNPS 2 steel pipe. From Tables C9.10a TP and C9.10b IP 5 Line length 180 ft (55 m) Two angle valves == 48 ft (146 m) Three elbows =_99t (Gom Total equivalent length = 237.9% (72.6 m) From the receiver to the evaporator, check the equivalen’ temperature drop in NPS? steel liquid line. Table C9.9 indicates a 1°? temperature drop per 100 feet of NPS 2 line (0.(2°K/m in DNSO.) :C504 HIPING SYSTEMS. ‘Temperature and pressure drop in NPS 2 liquid line for 10C tons (352 kPa in DNSO): At= 1.0 x B7.9/100 x (100/192)! = 0.662°F —_(0,368°C) Ap = 3.05 x 237.9/100 x (100/192) = 2.02 ps: (13.93 kPa) For the illustrated calculations, the selected line sizes are as folows: jot tine NPS 4 ‘DN 100) Discharge line NPS3 ‘DN &) Liquidline to recever NPS3 DN ®) Liquidline to expmsion valve NPS2_ ‘DN 50) The compressor selection must be adjusted for the ction anc discharge line loses, so the compresscr should be selected for the folbwing gauge pressures: Suction pressure = 46.51 ~ 445 =42.09 psig (290 xPa) Discharge pressure = 2226 + 568 ~2282 psig (1575KPa) ‘Secondary Coolants | . . aa Piging for secondary coo.ant service may bé designed by the methodsuséd for water. Consideration must be given to an economical choice of pressure loss and pump pover: In evaluation of pressure drop, the density and viscosity have a direct effect. ‘An addition’ considera‘ion is the compatibility of materials with the fluid. In the selection of the concent‘ation of water-based secondary coolants, it is desirable'to silect a mixture which has a freezing point at least 1810 20°F (0 to 11°C) below the operating temperatire expected iri the sygem, Use 0:lower freezing pohnt mixture: provides « larger margin to prevent freezing. However, this is costly because of feduced hea: transfer and increased pumpiag costs. The methocs in Chup. B8. may be applied for sizing of secondary coolant piping. Fo: four comnon water-based secondary coolants, Table C9.12 ia IP shows he specific gravity, freez~ ing point, and viscosity at various concentrations, Table 09.12 in SI shows percent by mass, density, specific heat, anc viscosity at a temperature 18°F (10°C) above the freezing point. This informaticn enables determination of the friction factor anc the subsequent pressure loss. OilReturn ‘The CFC and HCFC ani hycrocaibon refrigerants are niscible with mineral oil. ‘These refrigerants carry some oil in the liquid phase without separation, whereas thenewer HFC refrigerants require synthetic cil to provice mutual solubility. Some. practitioners suggest the use of strall additions of R-29 or other hydrocarbons to mprove tte oil circtlation when using ‘conventional mineral oils with HFC‘siwod sapere ¢ -auod durzcony exoge 1.02 1 sonodinis> v1 sonsoOHA L “Ap 12 19108 IiYA paredwes sonmord 200 - vee s os. ove- | wor © oy ost- | svt & ov op | seve oe a7 ce rent a ce oor | cart baa : fees toe ws gst | car x we : we 9% siz | swt w oz oz | «oot a 097 soz | sore ‘ wh ze, or oo wor « ota —atured —gAiwext | Ysoona guid © yawn? | [Kiso ‘aod, gfyand | -pAyooosiy awed —ghiwesd | ay Suzsoug — aurods Gureosy — oyjoads Boreary 2910s Buyzooig —oyseds | vonen soon 069 oupyédaag 02Ki8 suai, ‘2pLOTYD aiM}POS. PHOT wiTe §5. sojuadaig 1wejoog Aepuooas (dl) Z1'69 ATEVL 2.505e506 PIPING SYSTEMS TABLE C9.12a (SI) Secondary Coolant Properties of Calcium Chloride—Water Pecemt ty mass | Densiye | speeneneat | Dyn, viscosity Freezing point °C % gl? VkeK 10° Nvm* =5 90. 1078 3640 18% 10 140 nat 5340 2a 15 180 ust 3132 Bat 20 210 1200 2930 458 25 240 aa 268 62t 30 260 1258 780 831 -35 200 a az 122 40 295 26560 2660 14 45 310 Ba 2612 280 TABLE €9.12b (SI) Secondary Cooknt Properties of Sodiun Chleride—Water | bey) er eee -20 26 188 3310 420 * Propertcsat 100 sbovs the ficezingpoin. refrigerants. Some oil is carried with the discharge ges from most convenional vzpor compressors. In halocarbon direct-expansion syst:ms without oil separators, the oil will pass through the sysem, returning with the suction vapor. In fact, mineral oil n small anounts improves vaporizing heat transfer 2y reducing the surface tension of the >oiling liquid. Oil in greater quantities drastically reduces the vaporizing heat transfer in an eraporator. Reducing tte proportion of oil cizulat- ing in the system is accomplished oy use of synthetic oi by lower temperatures in the discharge gas to reduce the anount of vaporized ail, and by ‘he use of high- eficiency separating elements. 4 Oil discharged from the comprisso¥ is carrie through the discharge line 10 the candenser and through the liquid line to the receiver: Itthen proceeds through the licuid line fiom the receiver to the evaporator, ‘whert it will ecllect in ficoded exaporators as it does not vaporize. Various methods ace used to bleed a mixture of oil and liquid refrigerant from flooded evaporators into the suction line. Heat exchange belween the return suction line and the warm liquid feed line is a common method used to “dry” the returning refrigerant. The cooling of the liquid feed by the evaporaton of the returning rei7igerant conserves the refrigeration effect. Other msthods of cil return irvolve heatng the return mixture in a separate line thiough heat exchange with the discharge gas. Still other succesiful methods involve auto- mated refrigerant gas pumping devices such as ejector: or electrically heated oil stils that are intermittantly pressurized or oro equipped with pumps to trensfer istlled oil back into the suction of the compressor. After oil s introduced into the suction line by a direct expansion system or by ore of the just-noted methods from flooded evaporators, the gas veocity in vertical reumn risers must be maintained so that the oil will be carried up the riser The necessary velocities exzressed in tons (kW) capacity rave been experimentallyREFRIGERATION SYSTEMS PIPING C507 TABLEG9.12c (SI) Sccondary Coolant Properties of Ethylene Glyool—Water - Pereentmass | Densty* | Specific heat | Dyn, viscosity Freezing point °C % gla? Tikg:K 10 Nel? = 140 1003) 005 ‘aes -10 25 1035 3820 355 -13 305 1087 3650, 338 2 360 1058 30 790 25 10 1068 3340 124 0 455 107 3205 180 35 495 1084 3082 304 40 530 109% es 480 45 560 1095 2840 3200 TABLECS.12d Secondary Coolant Properties of Prorylene Glyed—Water Feroenimatt |” Beaiye | Seca | Dn yaonipe rea pit" ¢ ee | Pang | Noa = 3 70 7 oa an s ‘a ‘ors ts a 3 16 388 us 3 = os ine za 3 « 081 as aa ot s is 0. - 3 a is ao 10 2 ne) 2s mo Los 7 ion os mmo ™ Propertcs at 10° above freezing pont. . detemtined. Table C9.13 lists the minimum capacities ini copper tubes thet will carry oil up suction risers. Teble C9.14 lists tie minimun capacities in copper tubes thal will carry oil up discharge’ or hot gas risers. Mukipliers shown at the bottom of ‘Table 09.14 can be used for sizing oil return risers at various temperatures. ‘Wh:n a system operates,over a range of capacities, the siser sized ‘or the full load condition may not be functional ior oil retui at part Joads, Is such cates, a double riser will ncrease the range of loads that will provice oil retum. See Fig. 09.7 fer details of a double riser. Lines A and B cre sized so the two tegether will handle the full load with proper oil turn in accordance vith Tables C913 or 09.14, or with equivalent velocities of other refrigerants. Lint A should be sized for the minimum capacity. Since mineral oil has a very limited solubility with liquid ammonia, and is not mutually soluble vith vaporized ammoria, it does ot travel back to the compressor in dry tapor suction lines. Control of oi. in ammonia plants isusually accomplished ‘with efficient oil spparators, and manually operated oil drain 2ots, stills, and pump- out systems. Some degree of automaton is successfully used to retuen oil from evaporators to ammonia accumulators, such asis used in hot-gas defrosting methods, However, the separation of oll in the low side means that the principal method of oil con'Tol inammonia systems is provided by ol stils or oil pats at the accumulators in flooded o7 circulated system evaporstors. Additional automation of cil return iswoe ore U0 990 LEO TO TaD 90 oor oor rst cor yy aye 0 GLEO OHO 75D HAV DOP ose oe rate BD. ESO, perO HO. TOT ae scr etl ges ae 90 eed, Goro oro DOs oo wer ssl vel tS we oD 9e0. sO sero. sivO. DOE re cst or oes mv wd osc TO 6800" DOF Le e@ Oe tL SL wt 8ec0BBFO GD OTD 006 «OY coe ee TM EEL wt cco SBF 960 S8rO OL ew ce woe TEE se ort BLO Sev OOO eOD OS et ok sor gor sr St 090 coro, wo serO Lue eee ow = rot sor er aL 090. sero wwO serO os Sf cm 9M GOL TD STL 8090 coro HHO ETO OE sx oc cers tay red BO ED SoD TD. oso te TER Cd vie cel stl BOS 9860 sero BeEO 66K TIO oO rie cot yo 9 Ose Lod “eEO PSTD STO 9D]. OE OU Tie cst yor 99 ee veo’ 290 eETO STO SAO wT sic Sst 901 089 OBE 8sco REO TD STO LAO. OOI- eo Lm fos sts ne siso $6co sero tro 9000 oor oz To 9M dees ge od t6¢0 FTO LITO S800 oOI~ ror 9 zest owso 8620 _isrOo GIO. L800” ooe- usr cies ceo Ope Serv vero eweoeeO TO. a, Pc tdi edwin So se8—_uonons wy wo EE CE Garey snddes “1 ak) ‘Sios]y, wopong dn quowuyenug UO 309 suoy, wy AmpedeD wonerBujey uN (al EBD FTAVL e5080 v0 180 180 ¥60 OT at ST wT cos 190 £0 oso 180 $60 LOT ETE Oct 9T PET 080 SFO 680 460 860- OT OE -prT _urT@ a “cosa pu Za ono om or om Geo O8 om " . Sem, om seamesaate on pry s070 2 &.smesodnot prnbry ‘Roquadns pv aero: pry 3.05 Uo Pa om a. el 1 we az sz1 een amen wD eT __OK ow’ si go FOr Bo mt oct 9190 sOFD. ITO. BETO OL we TH SO POTD 00% = SIT 9070 TOYO HD SED ONS. Oy sm su It wer ss wl TOT 90 wed uD strO oe of wet OS HER tv sor gs tacn tee een zo, Ta So EL «BB OGY I $60 T6O Se LoD LTO UE Ne ca woz Gtk |r ser zag zI¥O uz sar 50 OO, om = «6SL_ GOL 959” SUF Tel UL0 660 HTD OTD BOD. OE st ss Ol OLB oi T9LO TED 6ST 80 OT eu stl eS ee sor 90. weo Ivo er wo0D oe ZL wu OER SLE fo 090 zIEO Loco se LODO eo TU eROS . BOE 6660 16s0 f0r0 oO TO ODOT oae— sel 96 6D Oe 6089 LO HD. ETD HD. SOD. Te! 6 9 ete 6D IYO LTO. STO Se0;D S00 OT- eu t6 Dye sev _uevo _oFrO sO HOD SOD _—0OE-_—- ts ex rie oO 88 es ot Or zo zeeO CO HTD 006 re ol owe re aoe ft cot aero HO zIZO TO om sie ow «est OL aS Oot «aT so EO aeTO TOS gs ose ca rR ee est 160 sO toro. ATO TO O08 sm Yer Te 8 ey 81 60 OBFO.«TIEO. GEO. TD O09 ex coz Tere are ort mT Ten cea zn eto te su HNL CETL:SCLSCYCOT_—SCCET:— TKO LY B9O TO HOD HL 9% OL oO SL Meh OTT SRO tro Luz ITO 500 OE ve eWeCGEL CGB GTS «= HZ:SSCBIT. «GHOSE seTO ITO UE 509wero ok s spore I9VSet S909 zoo rer GOUTE Sve It LIT SOET PLEO z SOBStE SPRL YrCLY ILLIP YCLYC CHSTL MOLL oC “TLEZ —9SET L080 OPO WET LACEEL OMSL LUSH LEST IWEL SUER WHY SOFT GPT Z9B0 OL. OT WSSTRT SOOSOL FUOOS OOD'EE SeH'6I TETOL Sur ere “I9RT 401 OSv0 HCO =o s @ SETI HSTHOT HES OPOKE OOZIC SESOE Ges'9 BLE SCOT SEIT 9190' 60> OT GOGOL OVESTT 9583 SBSE WEIL CLL WHY OLE IPC MOLL ELLD OED 0 HOSEL EER LISLE 80'S “LYE SESL ORY BENT “BBCI SIBO sero poeD' Ss “ @ Wrort oeoos SLE“ ENSC ENETI —HORL HERP FELT CCHIT ThBO Toso EO" S— Grriyl SLYS8 GUT ETLZ BLL EIT LVS U9" OST sevO zsO sO. ST- suvss sor GROTH BRST SBS GLY L608 SISO LOCO BD SI- o- @ L198 GRTIS TIVE LEST OHS. DESY LGVE “ELT ‘uOCO ZECO. LIED ELTO” — uady sey BPE TA TS ‘sotayeuoducs ou pyMby 0K 04 He orrosion,‘REFRIGERATION SYSTEMS PIPING css Fittings. The Pressure Piping Code permits the se of standard fittings, provided they are compatiole with the reirigerent or fiuid The stanéard ratings of forged steel flanges, fittings, and similar parts may be used for refrgerant service Bel-and-spiget fittings may be usec only for vater and crainage service. Couplings made of cast, malleable, or wrougtt iron may not be used on pipe containing flammable or toxic fluids. ‘Wrought-iron couplirgs are sutject to the same imitations in temperature, stress, and service which apply (0 cast-iron screwed fittings. Valves Cast-iron gate valves and plug cocks must not be used in liquid-refrigeran: lines unless consideration is given to ‘he expansion of liquid trapped in a space when the valve is closed. Several manufacturers make standa-d lines of refrigeratioa fittings which do no: fall into the classfication of ANSI Standards for forged-steel valves. These valves and fitings over long years of usage have gained acceptability anc are widely used and acceptable fo: refrigeration service o the degree recommended by tke manufac. turer. Other Factors, The Pressure Piping Code lists the followingdynamic effects which should be taken into account in the design of ref-igerant piping. 1. Imgact forces (including hydraulic shock) catsed by eitrer external or inter nal conditions, 2 The effect of vind loading on exposed piping. 3. Piping systems located in regions where earthquakes ae a factor are to be designated fora horizontal force in conformity vith the good engineer.ng practice using governmental data as a guide in determining the earthquake force. How: ever, this force is not to be consdered as acting concurrently with lateral wind force, 4. Piping shall be arranged and supported with consideration for vibretion. The Pressure Piping Code also cals attentior to the folowing weight effects which should be :aken into account in the design of piping: 4. Live loacs suen as tae weight of the fluid transported and snow and ice leads the latter will 3 encountered. If lov-temperature pipingis not insuated, there ‘can be a buildip of ice on the pipe even in high ambient temperatires. 2 Deed loads, consisting of the weight of the piping components and insulatior and other superimposed loads. 3. Tes: loads which consist of the weight of the test fluid in the pipe. REFERENCES: 1, ASERAE Safety Code for Mechanical Reftigeration ,ANSI/ ASHRAE-15, Anerican Soct ety cf Heating, Refrigerating and Air Conditioning Engineers, Inc, 791 Tullle Circle NE, Atlanta, GA 30229. 2. ASNE Code forPressure Piping, SectionB31.5 Refrigeration Piping, ASME International, Three Park Avewe, New York, NY 1016-5990.C532 PIPING SYSTEMS 3. ASME Boler and Pressure Vessel Code, Section VIII, Untied Pressure Vessels, ASME Internatioral. Three Park Avenue. New York, NY 10016-$990, 4, ASHRAE Handbook of Fundamentals, American Society of Heating, Refrigerating and ‘Air Concitoning Engineers, In., 1791 Tulle Circle NE, Atlante, GA 30329. 5. ASTM Stardard BS28.02, Standard Practice for Making Capillary Joirte by Soldering of Copper and Copper Aloy Tube and Fittings, American Society for Tesing Materials, 100, Barr Harber Drive, West Conshotocken, PA 19428, 6. ISO Mechanical Refrigerating Systems Used for Cooling ind Heatine—Safety Require- ments, ISC-5149, International Stadards Organization, Geneva, Switzrland. 7. WY. Richards, Refrigerant Vapor Line Sizing Not Dependent on Length, 1983 World Congress o! Refrigeration, International Institute of Refrige-ation, 177, Boulevard Malesh- cerbes, F-75017 Paris, F-ance. 8, W.V. Richerds, Practica Pipe Sizingfor Refrigerant Vepor Lnes, 1984 Annual Proceedings, Internatioral Institute cf Ammonia Refrigeration, 1200 19th 3t. NW, Suit: 300, Washington, DC 20036-2412. >. ITAR Piping Praaices Fandbook, Chap. 1, Pipe Sizing Procedures,” International Institute ‘of Ammona Refrigeration, 1200 1%th St: NW, Suite 300, Washington, DC 20036-1412.CHAPTER C10 HAZARDOUS PIPING SYSTEMS Ronald W. Haupt, P.E.* Senior Consultant Pressure Piping Engiseering Ascociates; Int. Foster City, Californix INTRODUCTION Recent ‘erious accidents! involving unwanted releate of hazardous materials in the process industries worldwide have focised attenton on the need. to 2rovide,a process by which such accident occurrences can be prevented. This chapter discusses one aspect of that 2rocess—the design, construction, operation, maintenance, and modification of ha:ardous piping systems. The organization generally follows the sequencz of piping design, construction, and operation discussed in grecter depth in other chapters of this book. However, this chapter focuses on those elzments of the design, construction, and operation of hazardous piping systems that are of particular importance to the owner, designer, and cperator. Some of the elements discussed herein have been safely neglected in the design of more benign piping, ‘but are mentioned here because itis felt they shoud be considered for hazardous piping. “his chapter does not discuss in eny depth the requirements or methods to perform process sefety management, crcept'to ncte clemens of piping systems which may be susceptible to piping failures and neec to be considered in any system designed to prevent fluid handling system failures Lastly, this chapter does not discuss piping systems intended to hande auclear power or auclear pover waste materiab. These piping systems are highy regulated, and prescribed reqairements are published elsevhere. "This chapter i a major sevision to the prevous (Sxth Elition) “Toxic and Hazardous Systems Piping” chapter ty Richard C. Getz, PE. Chet Piphg Enginecr, Jnited Engincets and Cons uctos Inc, Philadetptia Pennsylvania, In dition, Philip D. Funner, PE. of Sonsumers Exergy Corporation, Cover Michigan provided tnfommatior and recouendauinns oa welding, Ala D. Nince, FE. of A.D. Nance Associates Inc, Evans, Georgia provided valving znd corstructin suggestion: and editoria comments, Dr. Wilian E. Gale, PE. of Bundy, Gale & Shiels, Novato, Clltomia provded resource information fon hazardous mateias, end Robert E, Serb. PE, Consultant of Rehmond, Calfernia provided an overall editorial review and eorments. 0393cosa PING SYSTEMS ‘The Ame:ican Society of Mechanical Engineers, Cod» for Pressure Piping, 331.3, Process Pipiag* defines a hazardous fluid service as: ‘a fluid serviee in which the potential for pertonndl exposure is judged ‘o be significans and in waich & singe exposure to a ver? small quantity of a toxic fiuid, caused by leakage, can produce serious irreversible harm to persoas on breathing or bodily contact, even when prompt restorative meisures are iaken [emphases added].” The emptasized words in the definition are all qualitative, and vhat may consti- tute a hazard in one ccndition (stam in a confined spice, for example) may not constitute a Lazard under other circumstances (i.e., steam in an open field). Afluid that may be a hazard to the uninformed general public, if exposed to it, may not be considered a hazard if knowledgeable service personnel are handling it. The responsibility of what constitutes a hazard ultimately resides with the owner, and the designer of a hazardous fluid handling system in concert with the owner. The * responsibility for a hazardous fluid handling system does not end with its design and construciion, Responsibility extends beyond systen start-up ‘0 the owner to ‘oferate and maintain the system safely and pravide for the emergency conditions of hazardous fluid releases. Piping tyrically will only be pert, perhaps even a small part;-of a fluid tand- ling system. Pressure vessels, pumps, heat eachange:s, turbines, compressors, and other primary fluid handling 2quipment items may represent greater cepital investments and contain greater quantities of bazardcus fluids, out all'parts of the fluid handling system, includirg the piping, must be designed operated; and muintained s> as to limit 'the release of the hazardous contents to nonhazardous levels. . ~ ee Gere ‘It must beemphasized that during the service life of a plant (as righ as 40 years ‘ormore), only diligent s‘ewardship will ensure safe, reliasle operation of hazardous piping systems, This stewardship tegins at the eaily design phase, and continues through detail design in:o the construction, erection, anc test phase and operation. However, thisis only the beginning—a comprehensive monitoring and preventative muintenance program must be developed and impleniented to-easure a hazard- free system. And finalls, any modifications or repaits need to be >erformed with the same care and considerations required for the origintl design aad construction, amplified by the recogn'tion that hazardous residues mey be present ‘and reative atand adjacent to the interface be;ween the existing coastruction and any new or adiitional fabrications. Table-C1C2 at the end of this chapter summarizes many of the design and operation recommendations developed herein, TERMS USED IN THIS CHAPTER: Definitions Sone of the terms ised in reference to hazardous systems are defined below. Other common pipingzelated tems are defined in Chap. Ai of this book;HAZARDOUS PIPING SYSTEMS C535 however, the following definitions of owner an¢ designer are specife to this chapter. Designer. ‘The person responsible for the design decisions te be made regarding the hazardous piping design. It is important that the designer have corsiderable piping design experience and have access to relevant expertse in the hazardous materials being hardled and regarding possible interaction beween the hazardous materials and the materials of construciion. Professional engneering registration ‘would seem to be anecessity, complemeated by an understancing of the intentions of the various codes and standards whici may be wed in the design, construction, ‘operation, and maintenance of hazardous piping systems. Explosive. Extrenely rapid combustion of a material such that a higt-pressure ‘wave is formed andpropagated away from the source, often causing extreme damage to nearty objects. Flammability. The degree of susceptibiity to igniton or compustion of « material under specific environmental conditions Owner. The organization which, will cperate the constructed hazardous piping system. The owner is typically viewed by legal systems as having tke overall Tesponsbility for tre safe design, constuction, ani operation of ownec facilities ‘and, as such, has the primary responiibility for the identification of hazards. While the designer may have some shared resporsibility and provide the owner with expertise witt which to evaluate hazards, it i nonetheltss usually held that the final responsibilty for the identification of hazards lies with the owner. Further, it does aot seem plausible that hazardous piping would ever be consiructed in a speculative envionment,’ end if an éngineering constructon is purchased by ‘an organization net involved in its design, the new owner should closely evaluate piping systems which handle hazardous and potentially hazardous materials. Process hazard analysis. A comprehersive review designed to identify potential harards and to produce specific recommendations ‘hat will reduce the probability ofa hazard occurrng. Process safety mamgement. The iystenatic proéeis of design, operatios, mainte nance, emergency planning, and training by which the unwanted release of hazard- ous matzrials is soight to be prevented or mitigated. Reactiviy. The susceptibility of materbls to relesse energy 2ither by themselves or in combination with other materials. - Risk-baced inspection. A process by which the evaluation cf the risks of tature are prioritized for ‘he purpose of develeping a system inspection program to limit system failures and the adverse conseqtences of those failures. Safeguarding. Th» provision of protective and preventative measures that reduce, the probability of iazardous conditions occurring. Toxicity. The abiity of a chemical molzcule or conpound to produce irjury once it reaches a susceptible site in or on the body. Acronyms and Abbreviations 29CFRIVLO OSHA, “Process SafetyManagement of High'y Hazardo.s Chemi- vals; Explosives and Blasting Agents” AIChE American Institute of Chemical Engineerse536 NIPING SYSTEMS APL American Petroleum Institute ASME American Society of Mechanical Engireers BIOS ASME B65 “Pize Flanges and Flanged Fi:tings” BI6S ASMEBI69 “Fa:tory Made Wrought steel Buttvelding Fitangs”” Bl6.11 ASME B16.11 “Forged Fittings, Socket Welding nd Threaded” BB ASME B31 “‘Cod: for Pressure Piping" BULL ASME B31.1 Power Piping BS ASME B31.3 Process Piping BUS ASME B318 Gat Transmission and Distribution Piping Sysiems B?S Bondirg Procedure Specification GR Code cf Federal Regulations HAZOP Hazards and Operability Study MSS Manufacturers Standardization Society ofthe Valveand Fittings In- justly NACE Natiomil Association of Corrosion Engineers NBBPVI _ National Board o' Boiler and Pressure Vessel Inssectors NDE Nondestructive examination : NEPA National Fire Proection Association NOSH ‘National Institute of Occupational Safety and Health OSHA Oscupetional Safety and Health Admixistration R&I Risk Based Inspection PHA - Process Hazard Analysis POR Procedare Qualification Record PM Process Safety Menagement wes ‘Welding Procedure Specification CODES, STANDARDS, AND REGULATIONS: The latest edition and aidendum cf an applicable code or standardshould beused for the design and construction of hazardous piping. It may be asumed that the latest edition and addendum of nost codes is the clecrest interpretation of the requirements of that cole. The cofes and standards meationed in -his chapter are US. codes ard standarcs, but could be reasonably replaced by comparable foreign coles and stndards. New-constuction piping codes and standards set forth engineering require- ments, normally considered minimum, 1o assure the sale design and constru:tion of piping systems. Coces and stendards use simplified approactes to gair the wisest possible usage. Codes and standards typically caution the owner and designer that these docaments are not handbooks; that not all. sesvice condiions and environmental effects cen be known; and, further, that requiements cannot be written to evaluate all such conditions and effects. The owner and desgner are asked to use their experience and engineering judgment to meet the sifetyHAZARDOUS PPING SYSTEMS. C537 factors inherent or explicit in the code: and stanlards. This necessity is com- pounded when the possibility of a hazard exists with a breach of tie fluid- containing boundary. "At present therc arc fow codes orstancards whichaddress th» area of operations and maittenance oj fluid handling systens, much less hazardous piping systems. ‘There is a growing awareness of the problem as evidenced ty recent U.S. PSM regulations! and efforts by industry to devalop epproiches to operation and mainte- nance, such as new and proposed codes aad standarcs by the AIChE, API, ASME, and NBEPVL, However, specific guidanc: for component evaluation in tiese new ‘and proposed codesand standards islimited: rather these documents tend to outline procedural requirenents and often rely on new-construction rules to guide the operation and maimenance of fluid-handing systens. ‘The jurisdiction and the obligations o’ the varios parties vith regard to codes and standards are ciscussed at length in Chap. Ad. Piping codes andstandards do not list hazardous substances;therefore the owner and designer must identify such materials, However, codes and standards may assist in the identifcatior process. The following isa listng of codes and standards, or documerts which may loosely be described as a code or standard; that may assist the owner and designer to identify hazardous materials, to properly design and construc: hazardous: piping systems, and to evaluale the performance of existing piping systems . API £70, “Piping Inspection Code: Laspection, Repair, Altecation, andRerating of Inservice Piping Systems” APL RP 574, “Iaspection of Piping, Tubing, Vawés, and Fttings” API RP 750, “Management of Process Hazards” ‘ASME:B31-“'Code for Pressure Piping.” The ASME‘B31 Code for Pressure Piping consists of several sections: B31.1 Power Piping, B31.3 Process Piping, etc, (ee Chap. Ad). All the sections publishec as separate books, may have useful information relevant fo the design and construction of bazardcus piping systems for thei: particular applicatioa. B3L.3, however, is the only sestion that includes a chapter on hazardous piping systems. . ASME B31G, “Remaining Strength of Corroded ripe" ASME Boiler and Pressure Vessel Code, Section VIII, Divisions 1 and2, “Rules for Construction of Pressure Vessels” ~~ NBBPVI, “National Board Inspection Code” NIOSH, “Registry of Toxic Effects cf Chemica Substances” NFPA 30, “Flammable and Combustible Liquics Code” NFP& 49, “Hazardous Chemical Da'a” NEPA 325M, “Tire Hazard Properties of Flamnable Liquids, Gases, and Vola- tile Solids" ° NFPA 491M, “Manual of Hazardous Chemical Reactions” ‘The references also include several riscellaneoxs noncode or standard publica- tions which may asist the owner and designer in their efforts toward theidentifica~ tion of 1azardous fluids and meeting PSM requirenents.** In acdition, U.S. regulations have been published (1992) which requireestablish- ing pro:edures to ensure the integrity of systems handling hazardous materials.(epiydsoud uatospAty) aurydsong @puioiye tduoqaea payes ose) auassoug psp vofouphy ‘aprxorad vadoup Ap) pony WaBOspAty sneaphque runt wabuead puts watoupse, (Poe onaoei2g pripes oste) poe spuontt Kaoqo19 erdeozou uyauIosoI9192 poe auojyprag (ploz onzoeaxo10g poye ose) pve snoap(aue ‘pie suoKDOInkLL ‘sworsovosonyexoH, veg (wneIo payte> ose) pipe oungmns Summing, (arewz0,3) epfyapreni0y aye Aya (epprorad auoyey (Aino voy ayes ost2) apexosod auoyoy [hur snospayae Sun “Pr ‘auzep ca suvpsoncpp cyt apxond lompya aonaqrenpplnorod Hdasdorn apixo Oe SLMAT t 148redorg popes osie) auddosdoworg-¢ aprionyie euros spuonyenuad aurwosg 2pUOIND aUIBIONE sulmoug 2puonyEn UoroR, spyonrin uo10g 2bin2 (youscronp)igh (epupay qaoKy Geusty popes os Aunowey Jo APHOL YI BUNCH SOW LOD TTAVL C538{2ppo18 anys pate O51 (oprapdyue asngns (pina) apreogp snsing (2ouphy Kuountry) outang ‘=pUORBESDY EMILIO, -ourorge¢ paites oste) apt epson c -ogdsoya pares ost) apuomptxo sesomdson [>9PIN) KUCQIED TOSINT © oppmns nafoxp(ey Conernn poieo os) spuciia unaram intr ‘sua[Alaoe OFOPIC (Ga p08 aprondospty sawn aux apuorys [KypoWy puw muDidoro}4 >, pROIP eUEOIND 539c40 HPING SYSTEMS Inernational efforts folow similar paths. In the United States, 29CFR1910, App. 2 Ists 140 chemicals wits high toxicity and reactivity which OSHIA considers tighly hazardous above the concentrations and threshold the quantities Lsted there. The chemicals ony are listed in Table C101. For concentrations, threshold quaniities, and revisions to the list of chemicals in Table C10.1, the latest CFR docunent should be consulted, Note that not all of the materials listed are in a liquid form capable of being piped. IDENTIFICATION OF HAZARDOUS PIPING The identificition of the hazards associated with the materials'to be hardled by a system is the firsi and mos" important step in tre piping design process. Cemplicating the process are himan error and inadequate managemert of mechanital sjstems, which have teen said to be the largest causes of hazartous accidents. Various methods may be used to identify these hazarcs. The “Emer- geacy Planniag and Cormunity Rght-to-Know Act” passed by the U.S. Conzress in 1986 established a famework for community emeigency plarning progams to deal with chemical releases. Additional federal actions eventually led te the pwwlication' ct 29CFR1310 znd identification of the hazardous materials listed in Table C101. It would not be typical for the piping cesigner to be as invelved in community emergency plannins as the owner: However, the piping designer may be more involved in the facilily's mandatory PHA, if covered by 29CFR.910, or more informial_hazards assessment, if not covered by 29CFR1910. From the designer's involvement in these tasks, it may be.a-simple matter. to estanlish whether a pising system néeds to,be designed as hazirdous."If die designer is no: involved in any hazard analysis, it is-inciimbent‘us0n the ovner to ersure ‘that all relevant knowledge developed: from these. anayyses is conmunicated to the designer, It is the’ responsibilty of both the desigier and the owner t be swe that all of the relevant information has been received and appropriately applied. Furtier, the designer is responsible to clearly communicate the design ingerit to the fabricator and erector and to provide the operstor with dear guidance as to the pioper opeiation of the- coristracted piping system’ A coriposite system viagmm (see Fg, Ci0.1)'may be useful in organizing pping system design features and in conmunicating them ‘to thé operstor. ‘Various methods are used to perform PHA. or less formal assessments such as HAZOPs (Hazard and Operability Studies), “whatif” checklists, FMEAs (Falure Mode and Effect Analyses), and “Fault-Tree Analyses.” These methods ofter can be used jointy, complementing cach other, to produce a higher cuality product. ‘While it is bevond the scope of this chapter to describe the differen: hazard assess- ment techniques, they all have sone basic objectives: Identify hazardous contents or finctions relating o safe handling of such contents Evaluate the operational and environmental effects aad their potential to cause a failure which would result in 1 hazard Evaluate the consequences of a postulated failure Develop svecific recommendations to contain, control, or limi: the identified hazard,HAZARDOUS ?IPING SYSTEMS C541 A INERT. CAS PLRGE colNeCTIONS PROF UNLOADING r possiblya rupture. Excessive deformation of a flange can also result ina leak but probably not ¢ ruptare.cae MPING SYSTEMS Dasign Conditions . The design pressures and temperatures should represeat the highest (or lovest if the piping system is to operate below ambient conditions) coincident pressures and temperatures the piping is expected to experience durirg any normal or abnormal operating condition, in:luding any possible environmeatal effects Possible varia- ‘tions in normal and abnormal opeiating conditions shoud not be alowed to exceed the design pressures: aad temperatures. The design pressures ard temperstures ard the maximum coincident extemal (hydraulic, mechanical, and environmental) loadings should be used to establish the design wall thicknesses and classes fr all fittings. Note that consdering the effect of external losds may necessitate higher chisses .of fitings than would be necessary for pressure-temperzture condtions alone. Based on variations in extemal loads, the classes of fittings nay vary vithin the piping system. Material Selection - : . Highly ductie materials should te used for hazerdous piping, The use of low- Svetility materials, such as cast-iron or glass, should be avoided. Materials with demonstrated brittle beravior ot sensitivity to thermal and mecharical shock oad- ings limit their serviceability and range of use. Low-ductility material are susceytible to brittle failare in hazard mitigation situations, for example. from. thermal shock when exposed to fire o: fire-fighting measures Imipact testing should be used to vecify material ductility, . Proprietary materials may be wed if the designer is confident cf thé'suitability ofthe materials for the range of normal and abnormal conditions the material will beexposed to, Materials: without iufficient service experience shotld De avoiled. ‘Material sclection should take iato account the suitatiity of the piping to resist deeriorationin service. Information should be sought to determine material perfor- mince in corrosion or erosion environments. Information may be provided; ether qualitatively (e.g., acceptable or not acceptable) or quantitatively ‘e.g, expressed + asa uniform corrosion rive). Quantitative information is ased to establish corrosion ‘orerdsion allowances: qualitative information to select a material thet will minimize eresion/corrosion damaze. Obviouly, the qualitative approach is preferable. » ‘The designer shouldbe aware that in actual service corrosion aid erosion will typically be localized and not occur uniformly throughou: the pipingsystem. Carro- sion is @ function of many parameters, such as trace elements in the material, temperature, flow geometry, and flow rate. Corrosion is usually accelerated at crevices, under backing rings, in threaded joints, in soccet-welded joints, at veld metal areas «nd heat-aifected zones, of in other stagnant, low-flow areas of the piping system: Erosion is typically accelerated where high flow velccities exist for example, in uadersized pipe, at elbows, branches, reducer, or other cations where there are flow disturbances. In addition to evaluating the siping base materials, the mettod of pip: or piping component manvfacture should be considered. Joints, materal postfomning treatment, and manufacuuring processes can render piping less resis ant to eroion/ corrosion than the base materials. Muallic Pipng. Metalic material identification for hezardous piving systens is ‘ou: of the scope of this chapter.. Numerous sources can provide meaningful Jata for use in the materials selection process” But the final choice of material shouldHAZARDOUS PIPING SYSTEMS 6.545 at least be reviewed by an expert in tte chemical and eroson and/or corrosion performance of the candidate material, ‘Metallic and wed material election should consider the potential adverse effects of any cissimilar materials that mey be in contact, as galvanit action at dissimilar joints can promote accelerated corrosion where it would not o:herwise be expected to occur. Material test reports should te sought for all pipe and components 30 they areavailable during maintenance activities to determine nore realistic factors: of safety against feilure. ‘Time-dependent material properties must also be addresseil during the material selection process. At high temperatures typically over 700 to 800° F (370 to 430° C), andover @ per.od of time ferrous materials lose ductility, decrease in strength, and flow (creep). Other metallic materials (¢.g., copper, alumiaum) expe-ience this phenomenon at lower temperatures and are typically not used where creep would affect performance. It may be wise to select a conservative design temperature when the hazardous piping system mus: operate in the creep regime. Additional supports and the judicious use of rigic supports can reduce pipe sag and pipe migration during service. . Nonmeultic Piping. Nonmetallic piping typically creeps at room temperature * and can suffer dezradation due to temperature effects or exposure to sunlight. Consideration should be given to additicnal supports and protection from sunlight. Using @ nonmetallic lining “in pipe or a metallic frame arcund the nonmetallic materia may offer protection or additional support. Providing protection for noame- tallics daring operation by isolation, dovble containment, or cther mean} may also ‘be appiopriaté Nonmetallics are attractive because of their co-rosion-resistant properties dnd low cost, but typicelly they are not erosion resistant. Also, nontietallics are susceptl- ble to undetectable damage due to impact during transport, storage, and irstalletion. Pipe-joining procedures require care that must be cbserved, bat because nonmetal- ics are perceived as “low tech," such procedures are often abridged. ‘Nonmetallics typically exhibit highly anisotropc behavior, Seeking the actual material properties (e.g, yield and tensie strengths, elastic rodulus) in the trans- verse ard longitudinal directions may be varranted'to assist in naintenanc: activities for the same reascn they, are sought for when usirg metallic materials: «=.» ‘Because of nonmetallic piping’s low clastic modulus, internal pressure can cause significsnt expansion of the pipe that i: similar to thermal expansion in metallic pipe. The performance of nonmetallics under cyclic loadings is not well understood because of the wide range of nonmetalit materids, many being proprietary, the Jack of material standards, and their typically norisotropic tehavior. Ii loads are expected to cycle nore than a couple o: hundred times, including pressare cycles, fatigue testing of the material may be varranted. ‘The designer should also be aware tat nonmecallics can often be combustible and may require fre protection. " ‘At this time (1999) there are 20 proven volunetric ND2 methods available. Visual and dye penetrant methods are used, but these methods are lim.ted to: the pipe surfaces, and usually to the outsid: surfaces. Ultrasonie methods have been suggested as offering a potential volumetric examination, bat these are thus far unproven. Refer t Chapter B13 end Part D of this handbock. Design Criteria Hazardous piping should be designed to a recognized pipiag code or standard. B31.3 has a chaptar dedicated to the design of hazardous piping, but other codescas PHING svsTEMs andstandards could be used with proper consideration of the hazardsand mitigating desgn attributes. Safety codes and standards, however, do-not adiress levels of hazard. The designer should conside: specifying greater sa‘ety attributes for hazard- ous piping to che degree he feels they may be necessary. A review of the 331.3 nomnandatory App. F, Precautionary Considerations, aid G, Saferuarding? aly provide the designer with idezs for enhancing the safety of the hazardous piping system. if the desiger plans (9 ute nonsandard or unlisted materials and components or to use more -igorous analysis or alfernative examinationmethods, tie justification for ‘he use of such materals and components under all tke possible operating ind envronmental conditions and the reasons for justifying the use of such analy and examination methods should b: fully documented for the owner's apprev, before the release of specifications and drawings for conitruction. PRESSURE DESIGN ‘The most significant safety attribute of piping is pressure containment. Hazardous piping should >¢ designed for the most severe, conditior of coincident pressure, temperature, and loading, B31 does not permit, nor shotid ary hazardous piping bbe lowed, to exceed design values, even for short periods of time, lt may, in fret, be wise to overdesign, the pipe for pressure (i.¢., specify. a.thicke> wall ‘than, is required by the piping coce used). Ocher loadings may be creased ty thicker pipe walb, but these other loadings, suck as thermal expansicn, may be compensaed for by adding te the length and location of pipe inthe ancher-io-ancher piping runs ogn soboatmeclab ok oye Stright Pipe | 6s S63 ‘The design of sraight:pipe should be in accordance with tae specified piping cote, Flov velocitiesshould be limited to a maximum of about 0 fUs (3 m/s) for liquids andabout 140 tus (43 m/s! for gases, especially where the potential for erosion and cormsion exists. Higher velocities mzy be justified by thorough investigation of the eroson/corrosion potential, but higher velocities may result in undesirable flew. induced vibrations and .ncise, i oe * . ~~ The additional wall thickness provided to allow for eroson and corrosion should be based not oily on the expected rate of wall loss and the anticipa‘ed life of the system, but also on the expected ecamination interval ‘of the in-service inspec. tion program. ‘Since hazarious piping is often small piping, protecion of the piping. from damage after iastallation is also advisable. An additional wall thickacss could be provided, or the piping nay be isolated or located in channels, cable trays, or double containments, Standard Components Standard piping componeats are discussed at length in this and other books (see Chap. A2), The designer, however, should be confident that components supplisd as standard actually meet the standerds specified; for example, the nanufacturerHAZARDOUS HPING SYSTEMS C547 should be able to provide évidence that tie B16.9 fitting furnished or a comparable fitting hes met the burst test provisions of B15.9. The designer should be careful to ensure that components are specified to match the pipe to which they are connected. Further, the designer should request to be advised ifthe fabricator or erector deviaies from matching components to the pipe Or uses « nonstandard fiting in place of a standard fing (eg, using a Schedule ‘160 fittig rather than Schedule 80 fitting, or replacng a standird welding tee with a febricaied branchconnection). Such deviations canresult in stiessesnot considered in the design. The designer should also be aware that velocities will inciease locally in non- straight oipe components over the mass (bulk) velocity in straight pipe and thus ‘caa inerease the lozal potential for erosion and coxrosion. Valves ‘A valve is usually defined by the piping codes as a standarc piping component. ‘As is the case with other standard components, the designer should be aware that fuid velocities, particularly in the vicinity of the velve seat, will be higher than in ‘matching pipe and can result in erosion and corrosion problens. ‘The selection of valve type, style, and valve bocy and trim materials should be based on hazardots fluid properties, required operating characteristics, and ease ‘of maintznance. When locating valves, the designer should consider required opera- bility and service access. Selection of valve materialsand location should alo include consideration of potential accidents (c.g, fire). The valve construction (ie., bolted bonnet rersus pressure seal or welded) should include consideration for pressure boundary integrity required to prevent leakage of hazardows material (see Fig. C102). itis recomnended that the valve meet recognized piping code requirements for design, materids, and manufacture. Isolation and Control Valves. The nuber aad locations of isslation valves should be based on system: requirements and include consideration for potential release of hhazardoas materiab in the event of pipe mupture or inadvertentimproper operation. Metd-seated gete valves should be of the flexible disc, split wedge, 3r double dise design which 2rovide tighter shut-off than soid wedge designs, ani prevent jamming of the gat, if the valve cools down in the dosed position. The wedge and/ or body seat shoul! be hard-faced to prevent galling. ~ Rempte-controled valves (with moto: operators air operators, etc.) versus man- ual valves should’be considered for more rapid valve operation and op:ration in acontaninated area. Similarly, remote manual operating devices, such asextension rods and chain-wheel operators, stould be considered for ease of and tafe valve manipulttion, For large valves where mo‘or operaters cannot te justified, sut where a valve may be recuired for isolation inan emergercy, consider using portable air drives to assist in rapid valve closure. Also, for large manual valves consider “im- pact” handwheels to assist in the seating or unseating of the disc. ‘Valve and operator type should be such that valve positon is easily visually determined (for example, see Fig. C12). For valves essertial to plant safety, consider the use of remote position indication. To preclude the possibility cf inadver- tent, imaroper valre operation, consideration shoud be giver to using valve posi- tion’locking deviees. Permissible valve leakage (i... leakage across ihe seat{s) when closed) should be based on the properties of the fluid media and hazard presented bv leakage.cs. PPING SYSTEMS Rising stom Valve sten FIGUREC102 Conventional gat valve, Valve backseat capabilityshould be considered for gland repacking while the sysiem is erergized. The design of valve gland injection or leakoffcollection devices should alsobe based cn the properties ofthe fluid media and hazerd presented by leakege. ‘Valve body drains should be considered for removal of fitid media fom the lover caviies for ease of maintenance activities. Control valve redundaacy (‘e., peralel installations) should be corsidered based ‘on sequized system operation. The designer should be aware that cavitation tan ‘ocetr in the piping downstream of control valves and can result in severe erosion! cornssion prob ems. Control valve positioner linkage materials need to be compati- ble with the external envronment ad contained fluids. Although service failures in valves have resulted from shrinkage cavities in castHAZARDOLS PIPING SYSTEMS ose valve walls,’ through-wall cracking of valves is rare, On the other hend, leakage ~ through the ste/bonnet is relatively common." Recent design enhancements have produced valves with sophisticated devices to eliminate valve stem leakage, Most of these designs employ a proprietary dellows-like device tc effect a hermetic seal ‘These devices, along with conventiona backup seals, may provide a solution to the problem of fugitve emissions emanating from velve stem areas. Reliefand Safety Valves, Relief and safety valves should conform to « recognized pressure comporent code which requires overpressure pro:ection of the pressury boundary. B31.3 requires that the desgn presswe shall not be exceeded by more than 1) pervent during a pressure-relieving event, Note tha: this is in conflict wit) the B21.3 requirement that piping shal be desigred for the most severe conditios of coircident pressure, temperature, and loading. 't is recommended thet the design pressure and temperature include vahies expected during the pressure-relieving event. However, should the 10 percent requirement of B31.3 te applied, che designer should consider including a temperature correction, if the design temperature is also exceeded during the pressure-relisving event. we Uness the pressure-relieving syste is designe! to regularly recycle or elimiriate the hazardous contents, the set pressures of relief and safely valves stould beset sufficiently above the normal and recoverable absormal operating coniitions stich that the pressurerelieving system is not activated except in an emergency. ‘Sinto small safety/reliet valves (Sentinel valves) have a iendeacy not to reseai fice they lift, they should have the spzing cap dhilled and tapped for a gag which should be attached to the valve to faciitate clostre. “stu Rupture disc and breaking pin devices are rormally unsuitable for relievitig pressure in hazardous piping. ro Nonetandard Components Specia fabricaticns, forging’, castings. and othe: specialties may cosiplicaié, the desigh and cause difficulties later, if replacements are, requed. However, if they should be qualified for pressure design it accordance with a recogrized piping code or standard prot test, such as foand in B169. The preof test procedure and the resilts shouldbe reviewed and approved by the designer. The proof est should prove he nonstandard component's cipability to function at 110 percent of the Piping system’s design pressire, adjusted for the design temperature, wita thé factor of safey inherent in the piping code wed for design. Expansion Joints and Flexible Hose Expansion joints and flexible hose shoald be avoided unless absolutely necessary to accommodate ‘hermal movements where expansion loops are not feasible or to ecouple the piping from potential sources of mezhanical vibration, (eg,, rotating equiprrent): If used, it should be realized that portions of the joint cr hose are ‘ypically highly stressed and can degraie rapidly in an adverse enviroament or if manufectured with matcrials unsuitabl: for the service. If te rods are not used, the pip: supports must be designed for :he very la:ge loads resulting from pressure thrust. The designer should follow the ‘ecommentations of she manufzcturer and the Expansion Jont Manufacturers Asiociation.!*C550 SIPING SYSTEMS Ws, whether welded or mechanical, must be suitable fo: the anticipated pressure, temperature, and extersal loading conditions. The designer should also ‘be aware thet joints may be sensitive to the type of loading or be subject to selective tack by the erosive o° corrosive nature of the fluid hendled. Bawriers and suiclds should be considered for mechancal joints to protect personnel from leakage. Welding Welding is the preferred method of making a joint in hazardous piping. Welding should be performed ia accordarce with the ASME Boiler and Pressure Vessel Code, Sec. IX, Welding and Brazing Qualifications. While welcs between pipe supports or other attacaments and the pressure-retainiag boundary should ako be ir accordance with Sec. IX, welds between pipe support items may not need to be ‘Sec. IX welés if they conform to an acknowledged structural welding code. Meeting the requirements of Sec. IX should assure the quality ofwelding. Section UX requires the prepara:ion of a wiitten welding procedure specification (WPS'. The procedure is qualified in a procedure qualification record (POR) Vhich docunents physical testing proving that. wells made in accordarce with tke WPS pravide the requirec mechanical propertes, essentially equivalent,to the base material properties. Just as important is the qualification of the welder, vhich is done to prove the atility of the welder tomake sound welds in accordance with the WPS. ‘Tais is also proven by testing samgle weldments done by the welde-. For haza:dous service, the designer may wish torequire more stringoat testing cf either or both the proceduie and welder and in-process checks to ensue the desived weld quality, ‘To mininize any adverse effecs from welding, the designer should review and approve the welding pracedures aad qualifications of the manufacturer, fabricator, ‘aad erector. The welding processes chosen should be bised on the materials >eing welded, the quality level desired, and the availability of rersonnel capable of d:pos- tng sound veld metal. The filler material should be selected to essentially natch the chemical composition and material properties of the base materials. Cormsive exvironmens, low temperatures (requiring toughness considerations), and high temperatures (requiring creep comiderations) may necessitate special attention to welding processes, Longitudinal welds used to manufacture pipe are typically made in accoréance with Sec. IX, but also typically these welds are penalized, using ¢ joint efficency factor to reduce the allcwable base material stress resultng in a thicker weld. Some cére should be exercised if welded pipe is to be used for hazardous piping B31 joint efficiency factors have beer subjectively developed and do not take into ‘account posible operating degracation (c.g., due to creep). If the joint efficency factors are less then 1.0, using B31 specified NDE can inprove the joint efficency factors to a value of 1.(, but this should not necessarily be construed as removing ‘the metallurgical or geometric discontinuity that the weldjointrepresents. Consider- ation should be given 1o using seamless pipe in hazarious piping. On the other hand, high-resolution NDE in conjunction with a mill hydrotest bas been shown taproduce ahigh-quality weld sean in welded pipe. Because snisotropic properties axe a typical :esult of the pipe mansfacturing process (with more disparity expected inseamless pipe), the designer maj consider testing for both transverse and lorgitu- inal, yield aad tensile p-operties asit is not unusual to have the maxmum properties given in material test reports.HAZARDOUS PIPING SYS"EMS. C551 Butt welding is also the preferred method of joining two pieces of pire because it typicilly produces a relatively small local siress riser in the form of surface imperfe:tions, compared to other types of welded joints, and a butt weld joint can be readily examined by most conventional noxdestructive techniques. Typically, with aazardous materials, « smooth irternal root condition is desired. In orde: to echieve this, the gas tungsien arc-welding (OTAW) process is often chosen, sometimes with the additional application of automatic welding. This is true because the GTAW process charcteristics allow for dose control of heat input. Other procssses may be used tc fill the weld joint alter the initial layers are depssited, Socket welds can be used if the designer evalustes and aczepts its deficiencies. Several problems could be anticipated vith the use of socket welds. Socket welds are susceptible to failures due to pressute cycling. The gaps required for fit-up and ‘erevicesin the soclet are also potential cebris collectors and places where corrosive elements cai accumulate: Recent improvements ix automatic welding ia the field renderssuch a reasonably vieble alternetive to soccet welding. Backing rings used for butt welds may also:be subject to problems smilar to the erevices in socket welds aad it may be wise to avoid their use. Branch connections are often made using propretary products (so-called “inte” grally reinforced” weld.on fittings). {t s important to obtain the manafacturer's recommendations for welding, in particular the atachment veld profiles, prior to their specification in the design. Its alsy important to specify a cover weld for the attachment weld to the header similar tothe requirements of the various B31 codes. If the branch connection is large, cover welds larger than these required by code ‘would be prudent Details of the required weld should be specified in the engi- neering design. ' TE-fatigue isa froblem, consideratiox shoald be given to grooming the surface of the finished circumferential and branch welds. For cleanliress, and ifthe inside of the pipe if accessible, grooming the irside surface of the weld might ako be con- sidered, Dissimilar metil welds should aot be used unless thé thermal expanson effects of the two materials are evaluated. If the dissimilar metal weld is: necessary; it should be a butt weld, not a socket wel! or a brarch connecion weld. : Brazing and Soldering Brazing and soldering should not be used in hazardous piping.” Bonding Itis obvious that the bonding of nonmstallie piping is criticel to the construction ofa safe hazardous piping system. However, ihe designer should recognize the fact that many bonds are carelessly made be:ause the we of plastis pipe is not typically treated with the sume regard as metallis piping. ‘The bonding o! nonmetallic piping component: should be done in accordance with a written bonding procedure specifcation (BFS). The BPS should be qualified by testing the mechanical properties of an assembly made in accordance with the BPS. Al bonders should be qualified by preparingan assemby in accordance with the BPS that is tested to prove the border can make acceptable bonds. Refer to Chaps. D1 and D? of this handbook.C52 PPING SYSTEMS Flanges It bears repesting that Janges shculd be designed for the maximum coincident pressure, temperature, and loading. This is a fact that ig often nissed beciuse typically flange standards list pressure and temperature ratings only. If piping loads aresignificant and if the flange is net properly sized, ther the flange will inevitably leak. Incroasisg the flange class (cg., using a Class 300 flange whee the presiure anc temperature would indicate the use of a Class 150 finge) will result in a joint Jess likely to leak and in flange belts that are less likely to be overstressed. The equivalent pressure method descrited in Design of Piping Systems! is suitable for sizing flanges with extemal loads and has been incorporated inte a numbe- of pipng codes. Flanges conforming ta B165 arerecommended. Flanges minufactured to other recognized cossensus standards may need to be used for large-diameter pipingand Would be acceptable with due consderation given 10 external load: and fit-up Standard flanges include flat face, raised face, tongue md groove, ring joint, and lap joint flanges. Gasket materials nclude rubber, cork, elastomeris compositon, spiral wound ind laminsted metallc-elastomer mixes, aad solid mtal rings. The effects of the hazardous contents 02 the gasket material shouid be considered “These flanges offer the advantages of ease of joint assembly, standardized dien- siors, and general availasility. A dkadvantage is that ascembly requires high bolt loading during initial seating to re:ain sufficient gasket pressure in service. This shoald be a ccnsideration if the joiat is subject to high cjclic ioads or operates in the creep regime of the flange matirials. Proprietary flanges ar: nonstandard components and should meet the require- ments described for nonitandard components, The designer or the manufactrer should also understand how the flarge will perform when subject toexternal Icad- ings These jonts typicaly have fine susface finishes ard-tight tokrances, T1ey usuilly require great cars in assomtly, and the tight gaps may create the potential for zrevice comosion. Published flange ratiags (pressire-temperature-material class) presume the proper selectien of gaske:ing and belting materials. The flange joint is an assembly of flange, gasket, and bolting. In mest cases the gasket seating surface is critica’ to proper joint scaling, Should the designer be inaltemive to the interdependency of the Jange assenbly components, the flanged joint may no: withstand the operating and environmental loadings. ‘The designer should endeavor to locate flanged joints where external loads are low. The loading on-a flanged joint includes internal pressure, bolt oads, and the piping operating and environmiental loads. ‘Weld-neck Langes are -ecommenied for hazardous piping systems Other flanzes may be used f due consideration is given to their detciencies. Other flanges, excepting lap jpint flanges, are typically joined to the pipe with a fillet weld, which is less desirable than a brit weld. The attachment (etrengh) fillet weld io typically mor: highly stressed and subject to a stress amplification éue to its geometry. Sip- on fanges, if used, shoud be double-welded with the seal (or back) weld & a minimum distance from the face of the flange. Care must »e taken te avoid having the seal weld interfere with the proper function of the ange. Threaded flanzes should be avoded duc to their susceptibility to crevice corrosion and becaise threads represent significant siress -isers. Socket welding flanges should only be usec for NPS 2(DN 50) and smaller When using socket velding flarges care mist be wed regarding fit-up te ensure a ninimum gap between the pipe aid the bottom of the socket, Hate flanges (flanges vithout hubs) should te avoided unless doutle-HAZARDOUSPIPING SYSTEMS 0.553 welded and the flange thickness is increased to offset the high attachment weld stresses Lap joint flanges are susceptibl to crevice corrosion and potential sealing problens. Crevices in any of the flanges may collec: contaminttes, and the collected ‘contaminates could present aproblem wien serviciag flanged piping and equipment. Flanged joints in hazardous fluid service can be safeguarded with flaage shields to confine any escaping fluid, ‘Threaced Connections ‘Threaded connections should be evoided, if possitle. Their vulnerability to fatigue damage is significant, especially where exposed tareads-are subject 0 corrosion. Note that meteriak susceptible to corrosion will corrode if exposed to unconditioned air. If sed then, external piping loads should also be kept as low as possible. Seal welds covering the exposed threads may reduce external damage, but exposed thread: within piging may also pose erosion, corrosion, and fatigue problems. Con- structicn documeats should specify seal welds whare approrriate, ‘Other Mechanical Joints ‘There are a number of proprietary mechanical jcints which may be acceptable if consideration is given to their deficiences. Generally these nechenical joints have crevices which cau trap fuids and would be susceptible to crevice corrosion. These proprietary joints should be treated rovided as close to the vessel as possizle. External Loads Analyses The evaluation of sustained stresses due to weight is normally less of a concern than limiting the displacements (sag) between supports. Limi:ing the sag between supports to about / in (3 mm) will limit weight stresses and, in vapor systems, the pooling of condensibles. Long vertical iisers or dcwncomers may requice vertical supports to limit weight stresses not typically calcated using piping code compli- ance computer programs Shor:-term sustained load stresses, such as single-phase flov pressure -ransients, ‘two-phese flow transients, wind, and earthquake, siould be evaluated. In general, limiting the displacements due to these short-term badings wil limit thei: potential to cause a piping collapse or rupture and likewise limit the fatgue damage, if these loadings are frequent. TWo-phase flow transient ‘orces may also be reduced by reducing flow velocities. Piping displacements car be limited by active supports, such as hydraulic and mechanical snubbers and sway braces, or passive supports, such as zapped framing around the pipe. Passive supports.are sreferred ever active supports because active supports typicaly require periodic maintenance. Gapped supports and the piping may each suffer impact loads durng these short-term loadings. Passive protection for the piping can be provided ay special coverings, sleeves surroundirg the insulation, dev.ces or materials designed to tate impact, heavier pipe walls, or pads in the viciniy of the: gepped suprort. ‘The startup-shutdown thermal stresstange (ustally based on normal operating conditicns, rather than design conditions) is typically analyzed to assure adequate piping flexibility. But all temperature renges, whether due to normal or abnormal ‘excursions, should be considered and evaluated aong with the startupshutdown range, if significant. Further, all other load ranges, such as due to pressure and short-term bydratlic and mechanical excursions, result in some cyclic (fatigue) damage. However, if an excursion doss not oftex occur ani the stress-range of the excursion is less than half the allowable startup-shutdowa stress-range, it will probably not resuk in significant fatigue damage. Remember ‘hat because of a low elastic modulus, noametallics will experence signifcant pressire expansion effects similar to thermal expension, often recuiring comideration of their stertup-shut- down stress-range: Pressure expansion will occur in metallic piping too, but it is usually only a small fraction of coincident thermal expansion. “The designer siould be aware that significant ‘atigue danage can result from relative anchor motions from any caus? and nee¢ to be corsidered in the cyclic (fatigue) analysis. The greatest number cf piping fatures due to earthquake motions have been caused by excessive relative anchor moiions. The 1umber of times that any significant cydic stress-range occur: must be estimated to properly evaluate a piping system for fatigue.C586 PIFING SYSTEMS Creep damige from external loads should only ozcur trom long-term sustained loacings. Using the desig temperature to determine the allowable stress for 1us- tained loads stould limit external Icad creep damage. Evaluation of local stresses may be necessary as the result of rapid fluid tempera~ ture changes at geometric discontinuities and at a material diseontinvity due to any temperature change. ‘The magnitudes of flow and mechanically induced vibretions are hard to predict, although they should be expected if flow rates are high or inthe vicinity of reciprocat- ing and rotating machinery. If expected, a conservative estimate of their magnitules maybe necesscry for analysis or thedesign should incorperate (eatuics to mitigate thei: effects. Vbration may only be realized after system cperation begins and may need to be evaluated then. Pipe Support Design ard Analysis Pipe supports are part of a piping system and should be of concern in hazardous piping becauseimproper upport design and their attachment to the pipe wall hive caused failures in piping systems.” For example, a support lug welded directly to the wall of a high-tomporsture pipe ander the insulation, vith tae otier end of she lug exposed to ‘he atmosphere can “sul” crack in the pressure boundary becaise Of the lack of lug flexibility near the pipe wall Attachment welds to and 2 porion of the attachment nearest the pipe vall should be considered as part ofthe pressure boundaryand evaluated ard construcred accerdingly. A distance xo less than 2r from the pressure boundary, where.” is the chickness of the pressure boundary, should be considered the “portion of ‘he attachment nesrest the pipe wall.” This means that within 2t materals should be pressure bounéary materials and welds should conform toSection IX. Beyond tat distence, support materidls should be suitable for the service. Note that B3l.t requires support materiak for metallic piping to be listed, but that this could be interpreted ashe attachment material should be listed forthe purpose of consider. ing the interaciion between the attachment and the pressare boundary. “The design locations of supports and the method of attachmert to the ppe should be reviewed with ihe piping designer, and to the bailding with the buildng Aesizner, for large piping, to assure that those locations aad the method of attach ment are acteptable, ‘Generally, sapports need not be iategrated with the analysis of thepiping system (.e., the support stiffness need not be incorporated in the analysis of the pipng system). Incorporating support stiffnesses in the piping analysis will ty>ically reduce loads on the stpports. THis should only be necessary if tie piping loads are very high However, supports should be stiffer than the pipe at the support point or they may not support the pipe. This should be able to be cssessd morely by observation by an experienced support designer. Fipe suppors should sipport the pipe in the operating condition. For this reason, pipesupports siould typically be adjustable and the piping system reviewed durng nornal operation and pipe supports adjusted as nezessaty. Pipe suoports should also be designed to facilitate sloping the pipe for drainage or venting. Nechanical expansion anchors in concrete should be avoided: cast in place or epory grouted or equivalent anchor: are preferable. ‘Threaded hangers should be designed so that the threads are xot subject to bending, .HAZARDOUS HPING SYSTEMS C57 OPERATING MANUAL PREPARATION After ccmpleting the design, the designer should prepare an operatinz manual which describes the normal and abnormel operating and environmental conditions assumedin the design of the hazardous piping, Reconmended operating and mainte- nance irstructions for active components should >e included. Manufacturers of active components,such as vaives, pumps, instrumeatation, and saubbers, will typi- cally provide operating and maintenance instructions. All design drawings, including ‘component manufacture:’s design drawirgs, should also be inchided, The operating manual should inchde eny other information the designer concludes is necessary to guide the operstor in operating and maintaining the hazardous piping system safely. Recall, as vas previously stated that a composite system diagram (see Fig. CIC1) may be useful to communicate the piping systems design features to the operator. FABRICATION AND ERECTION Fabrication and erection of hazardous piping shculd be pe-formed in as’ dean conditions as possible to avoid any coxtamination that would compromise the process that is to cocur with the fnishec system. Forming It woule be wise to verily that piping code required heat treatment is performed ‘on identified materials after cold or hot forming of components. Welding and Bonding . : ‘The welling and bonding procedures established ty the fabrisator's anc erector’s QC programs should be followed. The fabricator and erector may augment the required examinaton by additional spo: checking of the welling or banding be~ ing performed. Butt Welds. A butt weld is expected to fill the weld groove between two pieces of pipe from the inside surface to the outside surfice of the pipe. Any deviations from this practice, such as between two heavier thin nominal pipe fittings, should be reviewed with the designer. ‘The siping laycut and shop fabrications should consider tow to faciitate field welding. for example, a field weld should not be Iccated too near a wall such that the welder must ue a mirror or other cevice to finish the weld. Butt weld joints should be made witiout the we of backing rings aad inserts, unless ctherwise permitted by the engineering desgn. Branch Connection Welds. Shop welding of braach connections is preferred to field fabrication. Non-90° branches are more difficult to fabricate than 90°branches. ‘The engineering design may detail requied branch connections. If aot, itis reeom-C58 PIPING SYSTEMS ‘Cover wald orcates an fective radius t reduce stres from ‘extemal loid moments cn branch (an actus radius will ‘ther bsrennteniceincey aan Insite comer graund off creating ‘a constant veld leg wil sield a conctart transition) Giroumierential Seétion - FIGURE C102 Recommended branch connestion details, meéided that the fabricé'or or erector consider the following good practices to provide quality branch connections (see Fig. C103): ‘Thesharp insice comer oflarge high pressure branch connections shuld be ground ff, creating an effective inside radius. The outside welds.:o the header should aot have abrupt corners; cover welds, such as those specified in B31 codes, should be used to provide an effective radius between the branch éttachment weld and the header. The cover weld should effectively make a constant smootied transition around the circumference between the attachment weld and the header. For large brarch connedions, the cover weld should probably be larger than the minimum typizally required by the B31 codes. - Socket Welds. Unless otherwise specified in the engineerng design,a socket weld. is expected to be a 45° filet filling the joint created by inserting tke pipe into aHAZARDOUSPIPING SYSTEMS c.559 ‘Weld throu shoud bo greater than or equal to that of 45° fille; the veld should fl the “groove” crated by the machined surface ofthe fing and the sucfe ofthe pipe \ ‘Smoothig the weld te pipe transition will reduce stress ‘fom external load monents N aman Dy) Tee Socket welding fitting (eoproxinately 1-1/4) ok [ ‘Approxinately 1/16 inch before Lootion where ris welding (to minimize the ant contuinatis peossitility oferecking the an accumula ‘wed whea i cod) Pipe and socket veding * ‘iting should mateh (ee BI6.11) FIGURE C10.4 Recommended socket weld details. matchirg socket welding fitting (see Fig. C10.4). The weld is expected to fill the joint to the edge of the machined boss on the face of the fiting. This means the leg length (size) of the fillet weld is approximately 1% times the nominal matching. pipe thekness, Weld buildup on the unmachinec surface of the socket welding fiting & not expected. Piping codes may mention a preference for a concave (smoothed) weld, but such a weld shoud not reduce the thrcat of the veld below that of che 45° fillet expected. The concave smoothing shoulc only be done at the toe of tae weld, Bonding. The febricator and erector should tak: care that nonmetallic compo- nents and bonding materials are protected from damage o- deterioretion prior to assenbly. The ‘abricator and erectorshould be careful during fit-up that joints to be bonded are wel supported and that the bond i: allowed to cure for the time required by the BPS before ccnstruction supports are moved. Mechanical Joints Mecharical joints should be assembled in accordarce with the requirements of the fabricator or erector’s OC program.C560 TIPING SYSTEMS Faanges, Large flanges and theit gaskets need to be carefully aligned prior to belting. The torquing ofthe bolts needs to follow a sequence to achizve the required bat efficiency (ie, bolt tightness) without overloading the flange or overstressing the bolts, The gasketing, if not specified in the engineering design, needs 10 be reviewed witn the designer. Small flarges may not need to comply with the alignment requirements cf the piping codes. The flexitility of the attached small piping may be sufficient to not overstress the flange elements; hovever, this should be reviewed with the designer. Threaded Connections. Where the erector uscs threaded connections not s9¢ci- fied, but not prohibited. in the coastruction documents the érector should verify the acceptability of such. Where the erector finds threded connections spedfied in the construction documents but seal welds are not, the crector should consult the designer -egarding lis intentions. Otter Mechanical Joims, Fabricators and erectors siould make sure that all ‘mechanical jcints are properly assembled by following the manufacturer's recom- mendations, yhich typicelly describe'a methodto verify proper assembly. Docunen- taton of the verification may be warranted. Erection, The piping stould be erected in accordance with the engineering design and the piping code used in the design. Deviations from the design, requirec for constructabilivy should be reviewed with the designer, : - Bending o: forming requized fr fit-up or alignment may require postforning heut treatment in accordance with the piping design coce. Temporary supports for erectionshould not damage the piping of allow excesive deformations of the pipiag or equiament:to which the pping is attached. After completing the installation of the piping the original design drawings should be revised to shew all charges made; that is, the piping system’ drawings shauld be ‘‘asbuilt” and made par: of the tum-over documentation. EXAMINATION, INSPECTION, AND TESTING een, INSPECTION, AND TESTIN( Mott piping codes distinguish between examination anc inspection, examiniaion beiag the responsibility cf the manufacturer, fabricator, or erector, and inspecion being the responsibility of the owner. There is a tendency to limit examination ind inspection to the welds, bat more than the welds should b¢ looked at prior to piying system startup, especially for hazardous piping. . : The manuficturer an¢ erector cin use various NDE methods to minimize sub- stardard workmanship, and the owner can employ the seme or diffsrent methods to verify the manufacturer’s or erecior’s work. However, nost piping codes spesify the examinaticn (NDE) methods with little opportunity for variations. Along vith the prescribed methods, tiese same piping codes specify aeceptance sriteria apgro- prite to the type and extent of required examination. Coie inspection is normally limited to the awner re the manufacturer, fabricator, or exestor's work 10 the extent that the inspector is confident that the manufacturer, fabricctor, or erector has performed the proper examinations and has made repairs wher: appropriste. Hovever, the owner's inspection stould include an overall review to assure that the onstruction meets the engineering design?HAZARDOUS PIPING SYSTEMS cst Because the pping Is considered hazardous, the owner ot designer may wish te impose more rigorous examination requiremens (ie., citer NDE_methods or acceptance critera) than is required by the piping code being used. For example, the designer may want to limit the porosity in a but weld to less than thet permittee in a highly corrosive, high cyclic service. But nore rigorous or mutiple NDE methods may noi be required by the code being used and the acceptance criteriz of the codes is only appropriate for the required examination. This is true for al ASME pressure component codes ands illustrated by paraphrasing a 1985 B31.1 Interpretation.” Paraphrased Question: When an NDE method is not recuired by E311, but it used, must the acceptance criteria meet the requirements of B31.17 Paraphrosed Reply: The acceptance criteria are not withn the scope of B31, In addition, piping codes typically specify NDE methods that have teen proven over long periods of use. Thus, if morerigorous, alternate, o1 newer NDE method: are imposed by the owner or designer, then acceptance erite:ia appropriate for the NDE nethod ani extent of examination mast alio be deve oped and orovided in the engineering cesign, It should be ncted that the code exanination methods and their scceptance criteria do not correate well to specific failure moces. Defects detected by NDE methods are typically repaired because they are perceived rathe: han Plating the defect to a specific filure mectanism. The detection of defects having litle or nc bearing on piping conponent performancemay result in unneces- sary, aad sometines counterproductiverepairs." Fracture mechanics holds the prom: iso of ono day being able to correlate defects with failure modes, but that is no: necessarily a given in 1999. ‘Adapting the B313 in-process examination for nonmétallic piping may be prudest. ‘The owner's inspection should inclde a detailed review of the erected system prior to startup to ensure that the pipe components and sizes, support types ‘and orientation, valve types and orientation, and specialty items are in accordance with tne engineering design. The materials of constructior could be verified br spot checks or more rigorous forms of examination, if this was not already performed during construction. During startup the owner's inspecion should also include verfication that piping moves as is intended (e.g., spring hange: indiearors move to the hot location) and that no undue vibrations or movement: are observed. - “Extmination of a branch connection and socket weld joints is normally limited to vistal and surface examinations. However, with careful interpretation, radiogra- phy cm provide relevant information sbout these fabricatiors and, further, it might be exyected thatthe welding done could be of better quality if the marufacturer’s fabricator’s, or erector’s welders knew that some of their fabrications were going to be evalucted by radiography. Recallfrom the previous discussion that this exam nation may be beyond the requirements of the design piping code, and acceptance criteria would need to be based on 4 prior agreement between the owner and the contractor. Leak testing of the erected piping system is a normal requiremeat of pi codesand shoulc be performed as required by the piping code used in the design. ‘The required leac testing should be performed before the hézardous piping system is pairted or the insulation is applied so that the system welcs and other joints may be viewed as completed or assembled. Where a hyiirotest is impractical, s pneumati:C562 PING sysTEMS test or more rigorous examination may often be substitated. A sensitive lea} test? should be used to augment whatever other leak testing methods are used. If testing requirements are not specified in the engineering design, the erecter should review. wi the designer the expected leak-testing requiremens. A recent and very practical B31.3 code change pemnits exempting “closure” welds from hydrotesting if in-process and either radiography or ultrasonic exanina, ticns are sutstituted, This exemption is also useful for dealing with additions of smrall piping (e.g., instrumentation taps) to piping already tested. OPERATION AND MAINTENANCE ‘The operators should operate the piping system within the desigr parameters of the system. The operator should te trained to understand the des gn parameters, other possible events trat could cause a failure, and opereting conditions that can degrade the systen over time. The operator shculd be familiar witt the hazardous ‘material handling system in sufficient detal to be dble to identity signs of system distress should they appear. One possible source of this information is the plant hazards analysis study and the results of any continuing hazards analyses. If design pressures or temperstures exceed the design limits by more than 10 percent, consideration should >e given to having a enginee-ing evaluation of the excusion performed. The same consideration should be given if the piging system experiences an abnormal operation not considered in the original ign’ or is exposed to a severe environmental event (e.g., hurricane or major earthquake). it would be wise to maintein an operating log (time versus operating pressures and temperatures, including outage time) to be able to evaluste fatigue and creep damage. This need not be'a minute by minute log, but could log average values over a given period between identiied maximums and minimums, vith excursions beyond the maximums aid minimums documented, Based on an understandirg of the materials of construstion, the criginal designer coulc be consuked or an «ngi- neering study be used to devise such a log. Maintenance of the hazardous piping system shculd incluce establisiing an in-service inspectior program to detect system deterioration before such deterioration can cause a failure. The in-service inspection could be basec on emerging (in 1999) riskbased tecinology, tailored to the piping system, Risk- based technobgy seeks to rank the piping system components as tothe probatility of failure and the consequences cf failure and implement inspection base¢ on the ranking. The ranking is based on the component and the uid contents. component materials, geometry, and the history of like components; and fuid cortent’s chemistry; conditioning, flow, and temperature. Through-insulaion radography bas been successfully used in such a program to screen sections of @ piping sysiem for Iccal thinning, followed by ultasonic tesing to refine thickness meisurements, Other existing and developing NDE m:thods car be used for in-service moritoring, should they prove to be appropriate. Acoustic emssion is possibly oxe such appropriate developing method the acovsti emssion metiods endeiwvor to identify locations of high stress by “listening” for material grain-structure fracturing. As of 1999,several efforts are under Way to develop staxdards for fow evalua‘ion and inspection planning. These include: *HAZARDOUS HPING SYSTEMS e503 + B31.1 is revising its nonmandatory appendix “Recommendec Practice for Opera tion, Maintenance, and Modification of Power Pising Systems” and ecnsidering making it mandatory. + B31.8 and the APT are developing “fitness for service” rules + ASME and the API are developing RBI guideline: for the power and pe-rochemi- cal industries. + ASME Post-Construction Committee s developing new staadards for post-con- struction conditions. ‘These activities should be monitore! for the completion and publication of their efforts. ‘The operator should maintain maiatenance records of the system and its critical components. These records: should be of sufficient detail to provide information of any significant change ‘0 the physical strucure of the pipe or piping component:, Any changes in operating parameters should also be docu- mented, since changes'in. temperature, sressure, fow, or the amount cf cyciing, or the chemical composition of contents,can result in increased corrosior, erosion, decreased serviceability of components or 2 decrease in thr life expectancy of materials Piping which operates in the creep range oris subjected to high thermal expansion stresses will self-spring (or shake down). Caution must be exercised when unbolting flanges or cutting loose piping that has sdf-spring. Large movernents of the flanges or pipe ends may result, If no problems existed with the previous operation of near equipment, reassembly of the piping should bo done by pulling the flanges or pipe ends together and rebolting or rewelding, ‘The pipe supports may. offer considerable information a2out system perior- mince. Lift off or adjustments to maintein piping in its desigr location may be af indication of piping system drift or the results of system loads not considered in the original design. Spring supportsshould be monitored to see if operating between, the operating and ambient conditions moves the bad indica.or between the op- erating end ambieat marks on the sprirg hanger assembly; nonmovement would obviously indicate the spring is not functioning; variations from: the expected move- ments may indicate piping system drift or loads at variance with the predicted design loads. MODIHCATIONS AND REPAIRS Hazardcus piping system modifications should be made in accordance with the latest piping codes. The interaction between the nev piping ard the existng piping should be evaluated. Piping system drawings should >e revised io show themodifica- tions. Repairs should also be made in accordance with the latest piping codes. The effect that the repair will have on thé oiping system should be considered. The repair should be fully documented as 10 the cause necessitating the repair and whatever repairs arc made. Any revisions to the piping drawings neccesary to reflect the repair should be made (eg, due to erosion/corrosion damage a segment of pipe mght be repaired by a replacemert segment of a more corrosion- resistant material)cs6a >MIPING SYSTEMS TABLE C10.2 Design ard Operation Recommendations for Hazardous Piping Icentify and give hezardoss piping syitems special contderation. Braluate consequence of siping failure (for ‘example, quantities released, persoanel ‘exposure, harm te the environment). Understand operating modes of the system, including veriations in rormal and ibnor- mal operating conditions Eliminate or ninimize the nse of meciani- «cal joints. - Provide smooth tramsition: at welded oints, ‘Choose valvesto be consitent with bizard. bus service Provide desigrs 16 minimize fugitive emis sions, We appropiiae NDE meiticas i assure "quality fabreation‘and erection, ‘Try to prsvide advies to pant designers re- sarding piping layout needs. ‘Us piping geometry to compensate fer thermal expinsion and contraction. Provide a collection and disposal system for Pressure relist of hazardous systems Sesregate hazardous piping systems ding ‘abrication, erection, anc testing to Sacli= fate all requirements being met. ‘Datign and msintain supports as part of tbe piping sytem. Provide design details of citical elements ‘» construct. ‘Provide mechanism for positive identifca- ‘ion of piping materials of construction, Provide an in-ervice monioring progam ‘br early detection of problems. Miintain service records throughout lit of system, Peiiodically eximiné critic elements “DONT" ‘View hazardois piping sy:tems as comen- tional systens. Assume piping system carnot fail Expect operatng conditions to be without variation. Overlook potential dynamic effects. Disregard shon-term loadngs combined With sustained loads or the fatigue difects of short-tern loadings. Choose materials sensitive to corrosion or erosion. Use low-ductilty materials such as cast- iron or glass Use mechanical joinis without consideing, means to saleguard then. * Hive abrupt cianges in jomt geometry, Use stem packing designs ‘hat can leak, Forget to perform a sensitive loak test with an appropiicte sensitivit!, Limit.NDE methods to those iti codé¢and. standards if ewer methinds will give reli- able results: 2 he Believe that phint designer will uindéirtasid hazardous pbing layout needs. Use expansion joints. 6 Vent arcu to amospher without proper treatment, ‘Treat hazardéxs piping like other systems. ‘Treat piping supports as injependent com- ponents. Leave critical frbrication and assembly de- tails to be provided by field. Rely on specifying materias with no fa- low-up. Wait for a cataitrophic event. ‘Repair piping vithou: documenting it, Insiall system and forget atout itHAZARDOUS "IPING SYS"EMS C565 REFERENCES 1, Trevor A. Kletz What Went Wrong? Publshing Company, 1985. 2. Amezican Socicty of Mechanical Enginesrs, D313 Frocess Piping, Americas Society of Mectanical Engineers, NY, 1999 Editior, 3. Occupational Safety ond Health Administration, 2” CFR Part 1910, “Process Safety Mamigement of Highly Hazardous Chemicals; Explosives and Blasting Agents; Final Rule” 4, B.APatty, Indwirial Hygiene and Toxicology, Interscience Pubishers, Inc. 5, N. living Sax, Handbook of Dangerous Materials, Reinhold Pulishing Corp, NY. 6, Mark, Dennison, OSHA and EPA Process Safety Menagemeni Requirements, Prectical Guide for Compliance, Van Nostrand Reinhold, A division of International Thompson Pubfshing Inc, New York, NY, 1994 7. Helnut Thielsct, Defects & Failures in Pressure Vessels & Piping, Reinholc Publishing Corp, NY, 1968, 8. MM, Kellogg, Design of Piping Systems John Wiley & Sons, NY, 1967 9, Philip A. Schweitzer, Handbook of Corrcsion Resistast Piping, Industrial Press, NY, 1969. 10, Waler R. Payne, “Toxicology and Proc:ss Design,” Chemical Engineering, April 1978. LL. Michael J, Wallaze, “Controlling FugitiveE missions,” Chemical Esgineering, August 1979. 12, AS.Krisher, “Flant Integrity Programs" Chonical Engineering Progress, 186. 13, CG Ziu, Handdook of Double Containment Piping Systems, MeGraw-Fill, New York, NY, 1995. 14, National Assocuition of Corrosion Engineers, “Coresion Data Survey,” Houston, TX, 198s. 15. ARC, Markl, “Piping Flexibility Analysis,” ASME Paper 53-A51, America Society of Mechanical Engincers, NY, 1953. 16. Expansion JointManufacturers Associaton, “Standards of the Ezpansion Joint Maaufac- turers Association, Inc,” White Plains, NY, 1980. 11, American Society of Mechanical Engheers, “Intapretations No. 7 to ANS/ASME 311," Americin Society of Mechanicel Engineers, NY, 1985. Case Histores of Process Plant Disasters, GulfCHAPTER C11 SLURRY AND SLUDGE PIPING Ramesh L. Gandhi Principal Engineer Bechtel Corporation ‘San Fraxcisco, CA DEFINITION AND BACKGROUND Slurry is a mixture of solids and liquid. A sludge denotes a mid or a concentrated slurry having a considerable amount of fine material that imparts high viscosity. Typical examples of slurries are the sold-liquid mixtures encountered in mineral processing plants and dredged materia. from waterways-and dams. Most of the slurries are made up with water. However, industrial paints, rocket fuel, coal-oil mixture, and coal-methanol slurries are made up vith liquids other thax ‘water. River sedimentin the form of slurry appears to have been handled since ancient times. All ancien: civilizations arose on river banks. Maintenance of waterways requires periodic dredging which resuts in a sand and silt water slury. Today dredging represents the largest volume of solids handled ir slurry fomn. Slurry transport is also used for dam construction, Blatch? reported the first hydraulic test results for a sandwater slurry flowing throughNPS i (DN 25) pipe. Gregory; ("Brien andFolsom, aid Howard reported results of tests of clay, sand, and gravel slurries. The flow of muds ard sludges through pipes was first examined by Caldwell and Babbit The first large-scale ‘experimental program on the flow cf sluries through pipes was -eported by Durand? ‘The conelations proposed by Durand ind his covorkers serve as a baiis for the present-day desigr methods. Design of a slury piping system involves * Selection of pipe diameter + Estimate of friction loss and pumping requirements * Selection of pipe material, valves, anc fittings * Selection of pumps * Selecion of inst:uments and control system for safe and reliable operation Pipelines transporting liquids such ¢s oil and water can be operated at any velocity up to their design limits. In most slurry applications. a certain minimum velocity needs to be maintained, to keep solids ‘rom seitling out in 10rizontal cs67cs68 PIPING SYSTEMS sections of the pipe. The velocity below which particles tend to sette out and form a deposit in the pipe is called the deposition velocity. The pipe diameter should be selected such that the velocity in the pipeline is maintained above the depostion vebcity over the operating range af flow rates. The operating flow rate range is determined by the expected range of solids throughput aad slurry concentration. Solids throughput is defined as the weight of solds to be tansported per unit tme. It is normally expreseed in tons per tour (tons/h). The slurry concentration is expressed as the weight of solids per unit ‘Wweght of slury, or volume of solids per unit volume of slurry. The slurry concentration may be established by the requirments of the upstream or downstream processitg facilities. This is normally the case with in- plant piping. In the case of long-cistance pipelincs, it tecomes advantageous (0 adjust the sluy characteristics and concentration to reduce the costof the pipeline sysem. An economic study is perforned to select parameters acceptable to upstream an downstream plants while offering economies in pipeline construction and cper- ation. The deposition velociy and friction loss in a given sizepipe at a given concertra- tion depend upoi the slurry flow behavior. ‘The selectior of pipe material, vaives, fittings, and pamps depends upon tie velocity of flow, abrasivity of the slurry,and pumping presiurcs which are in tum governed by the slurry fow behavior. SLURRY FLOW BEHAVIOR Flow of slurry in pipes depends upcn the interaction between the sclids and liquid ‘as well as between the slurry'and the pipe. a Depending upon’ the velocity of flow, pipe diameter; solids ‘size-distribution, fluid properties,'and solids characteristics, four different flow conditions car be encountered in a horizontal or nearly horizontal pipeline.! These are homogencous flow, heterogeneous flo, intermedate regime, and saltaion regime. Homogensous Flow Homageneou: flow implies that the solid particles aré uniformly distributed acoss the pipeline cross section. Homogeneous flow, or a clos: approxination to i, is encountered in slurries of high concentrations and fine pasticle sizes. Slurries exiib- iting homogeneous flow properties do not tend to settle end form adeposit under flowing conditions. Typical examples of homogeneous slurries are sewage sludge, coa-water fuel, clays, drilling mud, paper pulp, titania, fine limestonz (cement cin feed slurry), thorium oxiie, and miny other finely ground material. Heterogeneous Flow In heterogeneous flow conditions, ‘here is a pronounced concentration gradient cress the pipeline cross section. Slurries at low concentration with rapidly setting (course particles) solids generally exxibit heterogeneous flow. Typical examplesare sand and gravel slurries, coarse coa. slurries, and coarse ‘ailings slu-ries.SLURRY AND SLUDGE PIPING ce Intermediate Ragime This type of flow occurs when some ofthe particlss are homogeneously distributed while others are heterogeneously distributed. Most industrial applicatons involve wide range of particle sizes. Intermediate regime of flow is expected with transpor tation of tailingsslurry from mineral processing plants and transportation of coal water slurries. Saltation Regime ‘The flid tu:bulexce may notbe sufficient to keep fastsettling rarticles insuspension ~ The particles travel by discontinuous jumps or roll along a siding or stationary bed on the pipe bottom. This type of iow wll occur with coarse sind and gravel slurries, IN-PLANT SYSTEMS — INPIANT SYSTEMS In-plaat systems generally involve horizontal, vertcal, and inclined sections of pipe. ‘The pipe lengths are generally short. 4 large number of bends, valves, and fittings may be present n such systems. The pressure bbsses due to bends, valves, and fittings may be a significant part of the ‘otal iricticn loss. Staic head due to change in pipe elevation may be'a significant part of the total pumping head reyuirements for in-olant systems. Typicel examples of in-plant system ere slurry preparation Plants, mineral teneficiation plants, and municpal and industrial waste treat- ment plants. Jn 4 mineral seneficiation plant, dfferent ty2es of sluries may be handled in the same plan. The sturry concentration as well as the particle size distribu- tion of the slurry may change as the mineral passes throtgh various grinding, separation, and setiling stages. Large variations ia slurry characteristics and flow rate may be encountered in the same section o! pipe owing to changes in ore characteristics or plant operatiors. The pipes stould be szed for these antici. pated variations, LONG-DISTANCE PIPELINES A number of slu:ry pipelines have been built to transport solid paricles. The ‘materials transpoxted include coal, limestone, kaolin clay, Chna clay, iron concen. trate, copper and nickel concentrates, shosphate concentraks, gold ore, fly ash, sludges and minzral tailings. Because of the relatively Iong length of these pipelines, pressure losses through bend: and fittings are not a significent part of the total friction loss. Pumping requirements should include changes in pipeline elevaticn which could be substantial in long-distance pipelires traversing rugged terrain, Beccuse of the relatively large investnent required for a leng-distanc: pipeline, it is goncrally advantagcous to adjust tie characteristics of the slurry fo suit the pipeline requirements. The slurry concentration, particle sze distribution, and throughput are generally controlled within relativay narrow operating limits. The70 PING SYSTEMS material is generally finely ground :o obtain a pseudohonogeneous flow condition in the pipeline, y Slurry pipeline systems range from single-station low-pressure centrifigal Pump installations to multistation high-pressure reciprocating punp systems, In all cases, the basic requirement for successful slurry pumfing is to meintain pipeline fiov above a minimum operating velocity. The minimum operating vel at desired margin of safety above the critical velocity. The critical velocity in ‘umn is determined by the solids screen analysis, solids density, and concextration as well as the specific system characteristic—pipe diameter, slurry temperature, ete. :n a positive displacement systen, the flow is controlled by Varying the pump spezd. This caa be accomplished by the use of a fluid ‘coupling. eddy current couping, ac or de variable-speed drive, or a hydraulic clutch system. Diesel driven pumps have also been used.in remote areas where electric power was not available. For optmum system efficiency, most pumps should be operated at their maxium desgn speed. A short listrepresentative of the ong-distance slurry pipelines in use throughout the world-is presented in Table C1i.1. FABLE 611.1. List of Seiected Long-bistinee Suny Pipsiines .Througnptt,,“Lengia, "start of _Naterjal .,...- System, location ©" mmtpy Bi, | .operaion Coal’, °° \"Céniolidaion Coal, Chio 12 10s 1957" Russia 40 7 1965 veer ot + France (Mérlebaeh) 0 os 1g Goce ast siti es «Black Mes, Arizona © 50-9 F273, 1) we Tapan 03 16 1965, Limestone. Trinidad 0s 6 195) Rugby, England 15 37. 1964 Calaveras, California 1a v7 gr. Gladstone, Austealia 20 5 198. Iron Savoge River, Tasmania 23 53 196" ,Poleentrate Pens Colo-ada, Mexicn 16 30 39% Las Truchis, Mexico 14 11 1975 Sierra Grande, Argentina 19 20 7 ceosen + Samarco, Brazil 120 247 197% Kudremukh, India 75 42 1980 La Perla, Mexico 45237 198: 4ron sand Waipipi, New Zealand 10 4 i Copter Bougainvile, Papua, 10 7 19m corcentrate New Gunea West Irian, Indonesia 03 0 Pinto Valley, Arizona 04 u Kennecott Chino, New Mexico 07 7 KBI. Turkey 40 Kennecott, Utah a Phosphate —Valep, Brasil 0 eotcentrate Chevron, Utah 94 Makon, India 7SLURRY ANDSLUDGE PPING cs71 SLURRY CHARACTERISTICS Slurries may be clsssified as sottling sussensions and nonsettling suspensions. Set- king supensions require turbulense to mainaia iadvidual particles in notion or in suspension. Witt finely divided solids, homogeneous flow could also be achieved for setting cuspersions in turbulent flow. Nonsettling suspensions, as the name ‘implies, do not settle, even under the n9-flow condition. ‘Mineral concentrates, tailings, and coal slurries require tu’bulence tc maintain particles in suspension. Digested sludge and coal-vater fuel slurries do not settle under satic conditions, ‘The flow characteristics of a scttling suspension are largoly governed by the settling velocity of solidsin it. The flow caracteristics of a nonsettling suspension are governed by its rheological characteristics and dersities, Mos: commercial slurries contain appreciabie amounts of finely divided solds that change the racological propertes of the suspending fluid, For these slurries, both the settling chazacteristics ofsolids and rheological properties and the densityof the slurry become important, Slurry Density ‘The density of a slurry is given by 100 orm (C11. Tm = G0 GT cen Pm where py density of suspending liquid. Ib/f® (kg.m?), ‘fm = density of mixture, Ib/tt (kg/m?) lensity of the solids, Ib/f (x of the slurry. Figure Cl1.4 slows the variation of the trarsition critical Reynolis number with the flow behavior index n. Note that the Reynolds number for a rseudoplastic slurry is given by Rep= 8pD"V2-" (Gta) (=) (cu)SLURRY AND S-UDGE PIPING C879 oy tones 1800 , ~ ‘000 010 020 030 Gao os ceo O70 040 08 100 1.10 120 120 140 180 Fi aber nn FIGURECIL¢ Lamiar-turbulenttransition Reynolds number (Re,) as a Lnction of flow behavior Indes, n fr pseudoplanic slurties, where D — p= cansity (Ib 10), K = oonsistency (bts'f), g, = 322 Ibs/(Ibitt). where K'= consistency index, Ibf-s"/f€ (Past) . > ‘Re, = Reynolds mimber for pseudeplastic slury n= flow beaavior index For a yield pseudoslastic slurry, the generalized Reynolds number corresponding to the transition critical velocity can be eitimated by using the ‘ollowing equations, proposed by Hanks and Ricks." * - = 8454 oy 4 ,yeeoretn 142m "1 Fr, +3" a3 (c1.14) ‘The value of x, which is the ratio of yield stress to wall shear stress, at the critical Reynolds number s obtained from the following equation: 3232. eye ra , = (Abe) tn) — Hey = a4 ayer A) "(cua ‘The Reynolds namber Re, in Eq, (C1l.14) is the same as thst for a pseudoplastic in Eq. (£11.13). The Hedstrom number He, is defined as folows: DY a2 Hep Fe (2) (c11.36)380 HIPING SYSTEMS For given slurry charadteristics and pipe diameter, the Hedstrom number is com- puted by using Eq, (C:1.16). This value of the Hedstrom number is used in Eq. (C1115) to compute the value o! x. This value is thea used in Sq. (C11.14) t© compute the critical Reynolds aumber. The correspording transition velodty is caculated by using Eq. (C1113). Figure CiL.3 also shows the varitionin eritical Roynelds numbe: with Hedstrom number for the yield pseudoplastis slurries. Fzample CIS. Estimate the laninar-turbulent transiion critical velocity for the slurry considered in Example C14 if the pipe inside diameter is 12 in and the solids specific gravity is 3.0. Solution. Given det Pipe ID, D = 12in(0.3048 m) =f ‘Yield stress of slurry = 0.125 Ib/ft? (6 Pa! Slurry consistency k = 0.003 Ibf - s"/ft? (0.144Pa - 1") Solids specific gravity = 3.0 Solids concentration Cy = 50 Computed results: Solids density pp = 3 x 62.41bm/ft? (3000 kg/m>) Liquid density o; = 62.4 Ibm/f? (1000 kg/m?) Substituting in Eq. (CL1.1), we get Slurry density 2q = 93.6 Ibm /ft5 (1500 kg/m?) Compute the Hedstiom number Heyp using Eg. (C116): Heyp =9.9 x 10° From Fig. CI13 find the critical Reynolds number corresponding to n = 03 and the computed value of -he Hedstrom number. Reyp = 8890 Next compute the laminar-turbalent transition velocty, using Eq. (C11.13): » = 3.58 ft/s(1.09 m/s) Friction Lost n Laminar Flow. For a homogeneous tow condition, the friction Joss in a pipeline is estimated by using the following equation: a heap G) 5 cuaSLURRY ANDSLUDGE PPING C581 ripe irside diameter, ft (m) pipe length, ft (m) locity of fiow, ft/s (m/s) [= friction factor 4 = friction loss, ft (m) of slurry head 3 = acceleration due to gravity, ft/s? (m/s?) ‘The friction factor for a newtonian slury is given by the following: 16, fax (Ci1.18) Re = Ye (cus) Nm8e where ¢= slurry censity, Tbm/f? (kg/m) ‘an = slurry viscosity Ibf-s/ft? (Pa-s) = 322 Ibm-ft/(bE's* (1 for SI mits) ‘The laminar-flow friction factor for Bingham plastic slurries ix given by fol. He Het . 167 Re * RE 3fPRee (1120) ‘The Rernolds number Re and the Hedst:om number He are given by Eqs. (C1.11) and (Ci1.12), respectively. The laminar-flew friction factor for a pseudopastic slurry is given >y Dodge and Mezner" as 6 Rep f (C1121) where Re, is the generalized Reynolds number given by Ea. (C11.13). The laminar- flow frietion factor for a slurry exhibiting yield pseudoplastic flow behavior is. given by 16 fray (c11.22) where the Reynolds number Re, is givea by Eq. (711.13). (=x? | 2x2) = orp PR 2: w= (1+ 3ny'(1 =x) [cs 142m +i (C11.23) ‘The value of the ratio x of the yield stress to wall shear stiess is obtaned from renatin(st) (2) emcxge PIPING SYSTEMS ‘Pow Function ibaie e288 i 00E44 “1.005, 1.005608 1006 1 NE 08 com humvee FICURE C1LS Lominar flow function ¥ as 2 function of Hedstrom number He, = [D* y(t 'K)>"Y/(0Cg,), tor yield pseuioplasic slures, where D'~ pipe IL (ft), V= wlosiy (8s), p = density (bit), X = consistency (IDEs, 5, yield stress MDUB, g. = 322 Irsi(bt-t0), Figure C11.5 shows a plot of yas ¢ function of the Hedstrom number for various values of the Jow behavior index Stainsby aad Chilton® have proposed a hybrid model incorporating yield pseu- Estimate the friction loss in a 12-in-ID (304.8-mm-ID) pipe at a velocity of ft/s (0.914 m/s) for the slurry considered in Example C11.5. Given datz: Pipe ID D =1t (0.3043 m) Slury density = 93.6 1bm/ft° (1500kg/m)? Velocity » = 3ft/s(0.914 m/s) K ifrom Example C11.5) = 0.003 Ibf- s" /?(0.144Pa. s") ty ifrom Example C11.3) = 0.12SIbf/ft? (6 Pa) n (rom Bxanple C115) = 0.7 8 = 32.2it/s' (9.8m/s)* He,» (from Example C11.5) = 9.9 x 10°SLURRY ANDSLUDGE PING c5a3 Using Eq. (C11.13), find Re, = 7023. The Rernolds number is tess than the critical Reynolds number. The flow is laminar. From Fig. C115, the value of is found to be 0.184. The ‘riction factor 6 f= 7 Ree = 0.0124 cti i fur Friction loss per unit length = = 2gD 6.92 » 107 ft/ft(m/m) Friction Loss in Turbulent Flow. The friction factor for newtonian liquids in a turbulent-flow regime is given by the Colebrook equation: L D 9.35D TFB 5 tO Atoe(1 + ae) where 6 = the roughness of the pipe, ft (m), Figure C11.6 shows the friction factor as a function of Reynolds number with the relativeroughness¢/D as a parameter. Hanks and Dada"® have developed.a friction facor-Reynolds number relation- ship for Bingham plastic liquids, This relationship was later modified by Hanks!” (c1.2s) ERE? 55 ative soctones Regma Monber Rett) FIGURECIL6 Fricion factor fas a function of Reynolds rumber for newtonian sluies, where = pipe ID (ft), V = velocity (1),r= density (h/C),p = viseosty (DF SIF), g, = 32.2 bs (IGE)C884 PIPING SYSTEMS. x0 etontstrt seo “se ore aceee ec oats . «eo br, die FIGURE CHL? ‘Pricion fectr fas «Sinton of Reynolds guste for Ding pase sities, pe aDVilrg), He = Dolce) wie D = pip ID (6) V loci 9, = desi bl Wield sess (IRD, gs = 382 IgOE). bared on furher analysis of their data, Figure CLL7 shows the ‘riction factor— Reynolds number relationship for Bingham plastics. Note that for Hedstrom num- beis greater than 500,000, there is a long-transition region between the laminar- tubulant trassition critical Reynolds number and the Reynolds number at which the friction fector curve for Bingham plastics intersects the newtonian curve. The fricion factor for Bingtam plastic in this transition regon is significantly greater then the new-6nian friction factor. : Dodge and Metzer'*carried out a semitheoretical anilysis of the turbulentflow ‘ef seudoplasicliqudsin smoot pipe, They propoted ihe flowing equation for the friction fector in tuibulant flov: : 4 vf Figure C118 shows a pt of friction factor as a function of Reynolds number for vaiious values of n based on Eq. (C11.26). Hanks an¢ Ricks" have used the concept of mixing leagth in developing a semi- theoretical reationship between friction factor and Reynolds number for turbulent, flow of peeudoplastics. This method was extended by Hanks for estimating turbu- lent-flow friction factors for yield pseudoplastic fiuids. [nterested readers should refer to the referenced article for further details.” Figure C1LO presens the fricton factor-Reynolds number relationship with Hedstrom nunber as a sarameter for selected values of flow behavior index n. The following friction factor-Reynolds number relationship, developed by Tor- = log (re; 2?) = 2 (C11.26) a‘SLURRY ANE SLUDGE PPING C585 ae [amma fr2-2] Fi jos] {2.0} fos] on feo 2 cn acm. ‘Ee 16.08 ‘00s sss tae Peprebte ee PappMOENETIN, FIGURE C118 Frition ‘actor f a¢ a funetion of Reyndlds number ‘or pseudoplistic shurries. Re 2 BDV-"l+ SuMKg) where D = pipe ID ( V = velocy (i) p= deasty (bY 15, ~ consistency bt ef0) 928 (b # OEE ance, is applicasle to turbulent flow of newtonian, pseudeplastic, and Bingham plasticas well as yield pseudoplastic fiuids in smcoth pipes. F 2.69 295+ 3 oga— 2 “Shoe ( SPE + (C1127) where gppiye Ree Ra (c11.28) Torrance has alsoextended his analysis to rough vall fully turbulent frietion factor for noanewtoniar fluids and obtained ihe fellowing relationship: 2+60-28 (C11.29) Wilon and Tomas” and Thomas ind Wilson? have used a mixinglength ap- proach for the yied pseudoplastic fluids Their model includes corrections to changecse6 HIPING SYSTEMS \ _&&eos a * eae ‘a9 ‘ata a8 ro epotte tanto Rep FIGURE CI1Se_ Friction hetor fas finctionof Reynolds number for sield prendoplasti surries for n = 07, He = [Dipr(o/X "7 (Rg. Re = 8oD'V'Ywl2 + Sn)]r(Kg), vhere D = pipe 1D @h, V = vetocty (4), = density (I), K = consistancy (Ibt-e/8),%)~ eld trons (WUE), (gC 32.2 Ib- st). ugh vasens {be thickness the viscous sublayer according othe rhedtogy and atérn to aesount for the central unsheared core. Their miethod has been shown to give fairly good agreement with experimental data Stainsby and Chilton haye given @ turbulent-flow friction fa:tor correlation bised on their model. Example CII.7._ Estimate the fretion loss in a 12-in-ID (03048-m-ID) pipe at a velocity of 5 ft/s (1.52 1/s) for the slurry considered in Example C115. Since the vacity of flow is greater than the ransition velocity (se Example C11.5), the flow will be turbulent. Find :he Reynolds number correspording to the given velocity, using the slurry rheological data gven in Example C115. Rep = 13,644 ‘The Hedstrom number from Example C115 is 9.9 x 10*. Using Fig. C11.9c find the friction fector = 0.008. Substituting the values of D, V, f and gin Eq. (Cl..17), find the hea¢ loss per init length of pipe: Heaé loss = 0.0124 ft (m) slurry/ft im) pipe Heterogeneous Flow Fhid turbulence is needed to maintain particles in suspersion or motion in a horizontal pipe carrying setting suspension. At low velocities, particles settle and form a stationary bed on the bottom of @ horizontal pipe.SLURRY ANDSLUDGE PPING 587 a Neeee ‘1006 02) 3 M vee Toe Tone Taos ey Fos Mumba, ep FIGURECIL9_Frétion actor fas a fuactionof Reynolds sumer for yeld pecudophsti slaries fot n = 05, He = (Dpr(/K *)(A'g9, Re = 8pL-V""(nsz + on"), woere L = pipe (€,¥ loc (1), = densi (i), K = consistency (I=), = eld sess (A, & = S22(Ib-S/(bE-E). Deposition Velocity. The velocity bebw which ved deposits form is called the deposition velocity. Operating the pipeline at or below the deposition velocity for prolonged time conld result in a pipelineblockage. “he minimum operating velocity ina slumy pipeline should be kept greater than the deposition velocity, 1o prevent pipeline blockages - A number of empirical correlations have been proposed for estimating the deposition velocity. For uniform-size particles, the Durand™ correlatior, given as, follows, is widely tsed: Vp = Fiv2gD —1) (C11.30) is wae of F, io Durant comeation canbe obsined fon Fs, C1110 In Es, ‘11. 36) D = pipe DD, ft (m) Yp = depesition velocity, ft/s (m/s) g = acceleration dus to gravity, ft/s? (m/*) $ = solids specific gravity Zandi ard Govatos,* Wasp et al.” and Graf et al."have proposed minor modifica- tions of Durand’s correlation based on additional jata,cs88 PIPING SYSTEMS to0e.00. aa rf "nee oe ‘ae ese se emcee, ep FIGURE C119¢_ Friction frtor fas a forction of Reynolds number for yield peesdoplastic shrries for x = 03, He = [D'pn(o/X)*"]i(K°, Re = 8pD"V[n/@ + 61)"1(Kg,), wtere D = pire 1D (f), V = valocit (ts), p = density (lb/ft), K = consbtency (Io-2/f°), 5 yild stress (DINE), B= 322 (b-s/bE- ft. Most industrial applications involve nonuniform-size particles: Pilot plant test Tesilts or prior experierce with sinilar material is generally used for estimeting dofosition velocity for nonuniformsize particles. Oroskar ard Turian® have developed a semiempirical correlation which can be used for nonuniform-size particles. Their correlation is « follo QL Y"g) ers Vo = 4 5CV(1-C o= {sera ven ze Zz (cu3l) where Vp = deposition velocity, ft/s (m/s) 2 ~ function of 7/Vp, as shown in Fig, C1111 d = mean diameter of partcles, ft (m) ‘ttling velocity of solid particle in slurry, ft's (mls) m = hindered setling velocty exponent as a fundion of particle Reynolds number, shewn in Fig. C11.12 'w, = settling Velocity of soli particle in clear water of infinite extent, ft/s (nis) 4 = viscosity of water, Ibf-s/f2 (Pa's)SLURRY AND SLUDGE PLING 589 cone Cd can] 0 iv | a Z i en 5 so oan, 2 & 3 a can. oon) 5 oo © 0 0 40 Particle Diameter, d mm FIGURE C110 Durand coefficient F,_as a function of particle size where Vo = Fr ‘Vigb(s—1), where ¥p = deposition velocity 1/3), D = pipe ID (I), : = 322 tus, s = solids specific gavity. ‘The setling velocity w, of a single partéle in a fitid is givenas 2 — Salles ~ ad) (c11.22) "e 3Copr drag coefficient d = particle diameter, ft (m) = acceleration duc to gravity, 32.2 ft/s? (9.81 m/s") b, = density of solid particle, Ibm/ft? (ke/n) density of liquid, Ybmift (kg/m?) settlirg velocity, ft/s (m/s)cs90 PPING SYSTEMS NM we FIGURE CLLIL Function z a5 a function of w/w, for use in Eq. (C1131), where w = hindered settling velocity of sold particle in slur (ft/s), and w, = settling velocity of single parte in suspending liquid (fs) ‘The drag costicient of spherical particles depends upon the particle Reyrolds nuziber Re... For Re, < 0.1, the d-ag coefficient, Cp is given by Cp = 24/Re, - (C1133) where doiwo 8c and 4 = visecsity of liquid, Ibf-s/ft?(Pa-s). The value of Cp equals 04 for a particle Reynolds number greater than 10)0. Figure C11.13 shows variation in the drag coefficient with particle Reynolds number for spherical particles. Use of Cp Re, instead of Cp enables determination of the particle Reynolds nunber froma plot of Cp Re,? versus Re,, shown in Fig. C1113. Agd¥(ps ~ 1) Burge Rew (C1134) CpRe? = (C1135)SLURRY AND SLUDGE PPING 591 2 01 a 1 0 1 tom ‘00 rw FIGURECIL12 | Hinderedsettling velocity exponent as afunction of guricle Reynlds number Re, = daw/(ug), where d = particle diameter ft), w, = paricle settling velocity deasty (D/it), u = vscosity of fuid (Ibsit), g = 322 Ib-sAIbt-). The drag coetficiert is larger for a nonspterical particle than fer a spherical particle ofthe same diameter. Experimental dataon settlingvelocity asa function of particle diameter may be tsed to establish the relationship between Cy and Re, ‘The settling velocity of a solid partide in slurry (hinderec settling velocity) is given by Ww = Well ~ Cy)” (C1136) where tae exponent m varies with the particle Eeynolds number, as shown in Fig. C1112. Wilson®™ has ceveloped a semitheoretical analysis for heterogeneous slurries. He has oresented 1omographs for estirrating the deposition velocity for particles larger than 0.15 mn in diameter. Exampl C118. Estimate the deposition velocity for a sand-water slursy in a 12- in (0,30¢8-m) ID pipeline. The sand patticle diameter is 0.2 mm, and i's specific gravity is 2.65, The slurry concentration is 31.9 percent solids Soluion, Usitg Eq, (Cll.1), we find the slurry density, p, = 77.9 b/ft (1248 kgin). Using Eq. (C116), we find the volume fraction of solids C, = 015. From592 PING SYSTEMS see secon Chet « 108 : 3 5 bee 3 aT " zB roaes § 2 3 : & It : set « — a vt ens FIGURE C1113 Drag costicient Cy of spherical particles settling in Vquid. Cy = Agd(ps— 'n)/ [BW.p}, where d = particle diameter (f), w, = setting velocity (ft/s), ¢ = seceleration di to arvity (322 fel), = particle dencity Ub/#), o, = Liquid density (Ib/R), Fig. C11.10, we find the value of F: equal to 1.3 based on a particle diameter equal 100.2 mm and C, equal to 0.15. In Eq, (C1130) we substitute the following values tocompute the depositon velocity Vp: F=13. D=14(0.3048n) . g=32.2ft/s?9,8m/s?). . S = 2.65 Vo = 13.4,t/s (4.08 m/s) Example-C1!.9,_ Estinate the deposition velocity for a coal-water shirry ‘having a mean partisle diameter equal to0.01 in (0.0254 mm) n a T2-in (0.3038-m) pi ‘Tie volume ‘raction of coal in the slurry is 4, and the specific gravity of coal is 14, The viscosity of water is 2.1 10°S Ibf-sift? (1.0 m-2a-s) ‘We are given d= 0.01 in (0.254mm) = 8.33 x 10™ ft(2.54 x 10m) p = 62.4 1b/+° (1000 kg/m?) (Ps = 87.36 lb/ft? (1400 kg/m’) w= 2.1 x 107 Ibf- s/f? (0.001 Pa) Bc = 32.2 Ibrr - ft/(Ibf - s?) (Lin SIunits)SCURRY AND SLUDGE HPING cst ‘Thus: CoR® =84.5 From Fig. C1113, Re, = 27. Using Eq: (C11.32, we find w, equal > 0.035 ft/s (0.010; m/s). Next we find the hindered settling velocity «xponont 7 from Fig C112 m = 2.8, Now we compute the hindered setting velocity, using Eq. (C1136) Ww = 0.0084 *t/s (0.0025 m/s) ‘The raio wiv, = 0.24. From Fig. CLLAL, we find ; equal to 0.98. Next ve compute the deposition velocity, using Eq. (C1131). Vp = 3.5 t/s (1.07 m/s) Friction Loss for Heterogeneous Flow-Horizontal Pipes. “he formula propose¢ by Durand,” Eq. (C1137), is widely used for graled solids. The formala is based on sand-and-gravel slurries with particle sizes renging fron 0.2 to 25 mm, pipe diameters from 38 to 580 mm, and solids concentrations up to60 percentby volume, LC wie As — Ded, rictior Loss for slurry, ft (mi water per ft (mn) rictior loss for water, ft (m, water per ft (m) at same velocity. rag coefficient of suspended solid particle settling in fluid of infinite extent 5 = specific gravity of solid particles Zardi and Govatos® concluded fron an examination of 2549 data points that Durand’s formule predicted the observed head losses fairly well once the saltation data were separated from the heterog-neous flov data. Tuan and Yuan* have divided the available data into heteroreneous ficw, saltation flow, and flow with stationary bed and have developed conelations axplicable to each individual type of fiow. Their comelations fit the availeble data better than Durand’s correlation. In Eq, (C1137), the difference 1 ~ iy represeats an increase in pressure drop due to the preserce of solids in the suirry. The effect of particle sizz on slurry pressure drop is accounted ior by the inclusion o! the drag oefficient Co. In nost indusirial applications, the particle size is not uniform, A mean value of Cp & developed in Eq. (C1137) to secount for actual size variations. Equation (C1137) can be vritten as follows: (C11.37) ois con® (11.38) “is isin Secs —— =81 | —— Cu; (C11.39) (01 au? cen iw where C,, = volume fraction of solids taving size d; d= particle size of ith fraction Co; = drag coefficient of particle having sizz d; ‘N = total number of size fractions into which given perticle size distribution is dividedca PIPING SYSTEMS TABLE C112 Tyler Screm Sizes Sieve desisnation Sieve opening | Nominal wire diameter | Tyler equivalent Alternate | mm | in (approx) | mm | n (approx. | cesignaton 424 in [ioze | aza | oo 0.2520 4 in frog | 400 | 630 0.2480 3s in |'905 | 350 | 608 0.2394 3 in | 761 | 300 | 580 02283 25 in | 40 | 250 © | 550 02365 a2in [ssa | 2a |sas oases 200 ia | 508 | 200 | Sos 0.1909 min | 453 | 175 [aes 01807 Min | 381 | 150 |459 oes min | 320 | 125 | 423 106 in | 269 | 108° | 390 04585 1050 100 in | 254 | 100 | 3.80 0.1496 26mm | ‘in 26 | oss | 350 01278 0.883 in BOmm | %in 390 | 0750 | 330 0.1399 0.742 51 160mm | %in 160 | 065 | 3.00 0.1181 0.624 BSmm |(s30in [ass | 0530 - [275 01083 0525 27mm | in yar} 0500» | 267 ocs1 12mm [tein | 12 | 0498. | 245 0.0865 oat in 951mm | % in ast | 037s |227 0.0894 0371 in 00 mm | Se in goo | 0312 | 207 0.0815 24 mesh 573mm |C265in | 673 | 0265 - [ier 0.0736 3 535 mm | % in 635 | 0250 |182 oon? 3% 3,66 mm | tio. 3%in| 5.66 | 0223 | 168 0.0661 4 £76 mm | No. 4 476 | ong? | 1s4 0.0606 5 4.00 mm | No. 5 400 | 0157 [137 0.0539 6 3.36 mm | No. 6 336 | 0132 [123 o0a84 7 283 mm | No. 7 2a3'| oi’ |130 0.0830 8 238 mm | No. 8 238 | 0.0937 | 1.00 0.0394 9 200mm] v0.10 | 200 | 0767 -|0900 | 0034 | io \68mm | No.12 | 168-| o0s61 josio | oasi9 -| 12 laimm [No.4 |- iar | o0sss [ores | o0ss | 14 119mm }¥o.16 | 119 | 0.0469 | 0.650 | 0025s | 16 \o0mm}xo.18 | 100 | 0034 |osso | ooze | 20 apm | Xo.20 | oss} 0.0331 |osio | ooai | 24 Torpm | ro2s | o7o7| oo27s }o4so | ov | 28 samm | ho.30 | 0595] O02 [oss | oie | 32 500 um | X0.25 | 050] ooisr joao | oo | 35 20pm | 80.40 | 0.420| 0.0165 (0290 | oolig | 42 asinm | xoas | o3s1| 0039 jor | oom | 48 2s'um | ¥0.50 | 0297} oor |o21s | coos | 60SLURRY AND SLUDGE HPING C598 TABLEC11.2 Tybr Screen Sizes (Contineed) Sieve designation Sieve opening | Nominal wire diameter | Tyler equivstent Standard | Alternate | mm | in (approx.) | mm | in (qpprox) | cesignation 250m | No.60 0250] ome | 0.130 conor | 6s 210 un | No.70 cao} omss | ose ono | 80 Tun | No.80 0177] o@70 | 011 00052 | 100 149 un | No.100 | 0149] 0.059 | o110 oss | 15 125 un | No.120 | 125] oma | oon 00036 | 150 1054 | No.140 | o10s| omar | oor 00030 | 170 884 | No.170 | 0088} 0.0035 | 00s 0002s | 200 74un |No.200 | 007} ons | 0033 ono | 250 un |No.230 | 0063} os | oo 02017 | 770 Sun | No.270 | 0053] o01 | 0057 ons | Bs un | No.325 | oom | omi7 | 000 oniz | 00 37 um | No-#00 | 0037] 00s | ooss 03010 Note that the particle size distribution is determined by screen analysis. Tyler mesh screen: are widely used. The diameter of openisgs for Tyler screens is given in Table 211.2. Example C11.10, Estimate the friction loss for 1 coal-water slurry in a 12-in-ID (0:304E-m-ID) pipe at a velocity of 8 fi/s (2.44 mis) based on the following data: Coal specific gravity vy Volum: fraction of coal in slurry 02 Pipe roughness 0.002 ix (0.05 mm) Viscosity of wate: 2A x 10-5 Ibf-s/ft (1 m-Pas) ‘The particle size fistribution is as follows: Diameter, in (nm) Weight % 024 (6.1) 10 0112 6.05) 40 (0406 (1.52) 40 0.03 (0.76) 10 Solstion. Find the drag coefficient of particles of individual size fraction, Particle diameter aft (m) CoResi Rei Cw 002 (0.0061) 17x1e 171004 001 (0.00303) 16 x IF 52054 0005 (0.00152) 1.43 x 10! 45087 00025 (0.00076) 229 x 19 36176C396 HIPING SYSTEMS Now, EGuCph = 0.269 Next we find the friction loss i, for water. The Reyrolds numter for waer is 2.38 x 10’. And selative roughness of pipe = 0.0002/12 = 0.000167. From Fig, C16, f = 000325. Next we find ya n= 4a = 0.0129 ft/ft ‘Using Eq. (C1139), we get itty iy 1.97 Thus, Friction loss forslurry i = (1+ 1.97)ip = 0.0383 ft water/ft Friction Loss for Saltstion Flow-Hortzonsal Pipes. At low velocities or with particles having large se:tling velocities, ow with a moviag bed or seltating parcicles may arise. . “ ‘The saltation flow regime is enzountered when N, isless than <0. wep M= DET (C140) Newitt et al. have developed the following formula tased on their experiments: Sh = 66s — 2 (cr.41) Cy (C1i.42) Friction Lost for Intermediate Regime. Most of the formulas forheterogezeous flow and saltation regime are applicable to uniform particles. For amixture o: two or more size fractions, it is necessary to determine an averege pirticle diameter for use in various formulas. Use ofa weighted-average diameter or drag coefficient ‘wes illustrated in the previous secion. For slurries containing finely zound particles, a better approach is to divide soid particles into a fraction that is carried in homogeneous flow and a fraction that is carried in hetercgeneous or saltation regimes. The friction losses for each frection are computed value of Cp for use in Eq. (C11.37) is 50. Viscosity of water = 1 m-Pa-s (2 x 17° TbE-s/ft). Use steel pipe with a roughness of 4.57 x 10 m (0.00015 ft), Solution. Assuming that the depcsition velocity varies as the square root 0! the pipe inside diameter [refer to Eq. (C1130)], estimate the pipe diemeter tha: will give a velocity greater than the deposition velocity at the minimum flow rate ‘of 1000 gpm (227.5 m'/h). Minimum flow = (000 gpm ‘227.5 mh) = 2.23 f/s (0.063 m?/s) Let D = pipe ID, ft (m) Deposition vebcity = 4D"? ft/s Let the velocity at the minimum flow rate be 1 fs (0.3 m/si above the estimated deposition velocity.c.600 HIPING SYSTEMS ‘Thus, 223 wont =4VD+1 D =0.79ft (0.241 m) ‘Try 10.75-in (273-mm) outside-diameter pipe with 0.25-n (6.4-mm) wall thickness ani 0.375-in (9.5-mm) thick rubber lining. Pipe ID = 10.75 — 2(0.25 + 0.375) — 5 in = 0.79 ft (0.241 m) ‘Tre computations of fristion loss end pumping head recuirement are summarized in Table C113, Pipe Wall Thickness seen ye ney ais a) Tie pipe wal ‘thickriés: must be sutfcént to withstand ‘tie Sxpdied fhaximiim pressure in the pipe and expected corrosion-erosion effects on the pipe wall daring the intended operating lifetime. Most pipelines are designed to have a service life ofat least 10 years. An :xample based on the ASME B31.11 Code; With’ ‘Gabance- ments based an typical service-corditions, follows: Poin fe ‘The pipe wall thickness is computed by using the folowing equation: D syeiiy tte C15) where 1 = pie. wall thickness, in.(nm)>; maximum design pressure in pipe, psi (Pa) ‘S = maximum Allowable design stress, psi (Pa) : ‘¢ = allowance for-corrosion-erosion, inch (mm) we ‘The maximum allowable pipe stress is given as follows, per, ASME B31.11: § =0.8E x specified minimuin yield strength of pipe where E =: ald joint factor, as sunmarized in Table Ct Example C112. Determine thewall thickness required at point 4 in the pipsline shown in Fig.C11.16, A 12-in (324-nm) steel pipe istised The pipe s manufactured in accordance with APISLX. The ninimum yield strengh of the pipe is 52,00 psi (3:8.5 MPa). The design hydrauli: gradient is shown in Fig, C1116. The design hhead at point A is 800 f (244 m). The siurry specific grevity is 1.5. The metal loss dus to corrosion-erosion is estimated to be 4 mils/yr (01 mm/yr), and the design life of the pipeline is 25 yr.SLURRY AND SLUDGE HPING co: Solution. Meximum design pressure at A Maximum design pressure at A = 300 t slurry _ 800 x 1.5 231 pai = 520 psi (3589 kPa) ‘Maximum allowable stress = 0.8 x minimum yield strength of pipe steel 0.8 x 52,000 psi 416C psi 286.8 MPa) Corrosion allowance € = 0.004 x 25 in = 0.1in (2.5 mn) Wall thickness ¢ = ee +e _ 4160(2.75) ~~ 2(4160) = 0.18in 4.53 mm) +01 ‘The next-higher commercially available pipe wall thickness is 0.186 in (4.78 mm). Corrosion-Erosion Control Ina slurry pipeline, metal loss is expected to be a result of corrosion with possible ‘erosion of the corrosion products taking place simultaneously. Under scme condi tions, mechanical sbrasion will play a pirt in prodicing the metal loss. Erosive wear (ibrasive) is governed by the size, shape, end angulaity of the solids, surry concentration, and velocity of flow. Ina slurry pipeline, these parame- ters are interdepesdent to some extent. For examole, use of large solids requires an increase in minum transportation velocity. It las been found that above some critical velocity, the abrasive wear increases as the cube of durry velocity. Wear also increases as tke size of the solid particles increases. Thus, by reduciag the size Of the sclids, the abrasive wear can be su>stantially -educed due to the combination of lower required velocity and reduction in wear due to smaller particle size. The effect of slurry corcentration on the abrasive weat is more complicated From experience, i has been found that the meta loss due to abrasion isinsignifi- cant if the velocity of flow is less then about 10 ft/s @ m/s). For long-distance slurry pipelines, velocities in the range of 4 to 6 ft/s (1.2 0 1.8 m/s) result in an optimum design fiom the standpoint of economics. Thus, whea possible, a particle size should be selec:ed so that the slurry is nearly tomogenecusly susperded at velocities of 4 106 Ms (12 to 18 mis), Corrosion can be controlled by passivating either the anodic or the cathodic reaction at the pip: wall. Elimination o' dissolved oxygen ard the adjustment of slurry pH can reduce the corrosion rate substantially. In most long-distance slurry(ato) taen [avo | ono | saoo | oro (wa) yr oye 30) 580) vonoNS ZL YES Beg wusg SLLOUU TsO seer peEOUU Fagens ay JO WON “LL 1 atid yssauysnioy 610000 610000. 6Ico0'0 6100010 adid jo ssouyfnos oaneioy “OL so+aseo | so+ acs | so+ase9 | so secs A 40 04 spIOUADY 6 AID ro rs oor coor CoA aia tona yp Kimeas uoaip | S0-HOOT So-HO0T ro a8S'6 ‘voaas'6 Bed) 211/39) “FOIVA Jo KysOOsIA “L (eno) ban | z50 a zo wr ovitt s9 ove 79 ena (9119) ‘ba an En sto! sco sco ‘20 SpHOs JO WOE] SUIMTOA “p smo |___ ove ove | ove oe “Bie ogioede patos Bose /MOL v6 x9 at 4 Gre) spy ‘ARORA Z wong | sy vee sev0 sev g (Seas) sity IEE OPA “T Sanu swnsn | swasn | omen Su wor] nore | oa | aon | mop UA : suonwindenag aneInKIT IS keMME CrLLO Taw, c.602OZ pur GL Jo wns wie ese 98 oot. Camys (Ww) y peay Suydumnd peo “12 ‘axe — san) oz oe oro ors ‘rons (on) 3 (PEON AHPIG “Oz 81 Due “ZI “@) 91 (2) 9LJ0 wns, we es xe ost ‘umys (ww) y $80} peay 10K. “6L spueq ¥ x (827)/A x F | eT Ter Lo oro (wi) ag ‘spuaq ino} yBnosyp sso PEF “ST OMA XCIA XE | 0 ee 00 TCO. | (yy SoMeA 9am) WBNON 801 PEL “LT 2a «0 zo zo £00. ‘Aczns (ut) “odie ponAA (9) wan) oe eve et 960 ‘arms (4) 1p ‘adid pervozu0Hy (2) ‘Tamms (wl) 9 SOF UONDLE OL oe oF t6 v6 (a) aed peanen 30 wae St ost Ost ise “sy (us) y ‘add jryvoquoy Jo W8UET “pL yoy se owes | cbZ0'0 ort0'0 £y200, orto'o (ex) y/Aamnys (ur) y sadid prone, (7) 8 +ds Gampsjs9%m Y= fan Leno, 6020'0 we00 5020'0. (a) yams (w) Gers) ba =a | oLc0o0 t60°0 oxc00 src |G) urate Gear aie romorrrarr (9) =Kaanjs Jo} sso] vonaea “EL 403,C504 HIPING SYSTEMS TABLE C11.4 Weld Join Factor Weld joint factor Specificatior Pipe mfd. Pipe nid. no, Pipe type tefore 1959 | after (959 ASTM A53 | Seamless 1.00 1.0 Electric resistome-welded ost 100 Furnace lap-welied 0.80 0.80 Furrace battowelded 0.60, ASTM A105 | Seanless 1.00 1.08 ASTM A 134 Elearic fusion (arc) welded, 0.80 0.80 siagle- or dowole-pass ASTM A135 | learie resiotanve-welded 08st 1.00 ASTM A 139 Eledric fusion-velded, 0.80 0.80 deuble-submerged are-welded 4STMA 155 | Elearictusion-reided 0.9 1.06 ASTM A 38: Electric fusion-welded,, = LOE deuble-sibme ged are-welded ASTM A 67: | Elosric fuion-nelded — 1.9¢ 2P1SL Seanless a 1.00 1.06 Elecricresstanco-welded ast 9c lecri fesh-welded 08st 19¢ Elecric iniuction-welded = 19¢ Submerged are-velded = 19¢ ‘Furmee lap-welded 0.80 08s Furnace butt-weded_ 0.60 0.6C * Definitions for the varius pipe types weld joints) are given in paragraph 11002 of ASME 831.11 Coie. "TA weld jolt factor of 1.) may be usel for eectric resistance.welied or electri: lash-welde! pipe manufactured pr to 1959 where 1. Pipe furnished under thilassifcation1as been subjected tosupplenentaltesteaid/orheat trestment as agreed fo by the purciaser, and sud supplemental tesis andorheat treatment demonstraes the strength claracteristic ofthe weld tbe equal to the miaimum tensile stzength specified tr the pipet or 2, Pipe has been tested as required for s new pipeline ip accordanee with parapaph 1137.41 of the ASME B3i.11 Code. {For clases ind grades that have been tydrostatically and nondestnctvelyteetec to specification re- auiements {| Manufacture was dscontnued end process deleted from API SL ix 1982, pirelines carrying mineral concent-ate, the slurry pH is adjusted to 9.0 or higher, using lime to reduce the corrosion rate. Sleek Flow end Its Control Slack flow occurs when the availasle static head between the discharge point of the pipeline and a givea point in the pipeline exceeds the friction loss betveenSLURRY AND SLUDGE PIHING C.60S ou nah a wns Xx Batch hydraulic gradient without slackiow Pa Sa gin cE J 04] i 8 : N a vl a oe =a] | Kali 1000 T | EU fi I 3 Grund profile S 1.50.0) - ‘ ‘Disance, km FIGURECULI6. Piptine profile ind hydraulic gradient showing design ‘and slack flow area, gradient, bach gradient, those points at a given flow rate. It usualy oocurs jast downstream of peaks where the hydraulic gradient intersects the ground profile. For example, Fig. C11.16 shows slack flew area. In slack flow areas, the velocity of fow is gorcrned by velocity in the slack flow area is higher than that in the fully the pipe dope. The packed pire section. ‘The section of pipe which flows full ani is under pressure is called fully packed. Inslack flow secticns the pipe flows partly full, and pressure generally drops below atmospheric. Erosion takes place in the sack fiow area owing t io higher flow velocity ‘as well as possible cavitation at the point where te pipeline changes from slack flow to a packed flow condition Because of the aigher density of slurty compared to water, slack flow can occur when water is displaced by slurry. For example, Derammeliere and Chapman"* have described the development of slack flow in the Samarco iron concentrate pipeline during operation of the pipeline in batch mode. This is illustrated in Fig,e806 HIPING SYSTEMS CLL.16, where the hydraulic gradents for pipe full of slurry as vell as for batch operation are shown, Slack flow can be avoided by sing a smaller-diameter pipe section or crifice chokes. The velocity of low in the smaller-diameter pige should not exceed about AC ft/s (9 més), otherwise, erosion nay occur. Either orifee chokes or a combiration of orifice chokes and smaller-diameter pipe may be used to achieve flexbility ard econo. Storage Itis very sekiom that -he flow of solids in a slurry pipeline matches its rete of production at the mine site or itt rate of utilization at the terminal facility. To provide meaas by which the pipeline can operate efficiently by transporting solids ina nearly continuous ‘ashion, storage facilities are required at both ends cf the line. The type and amount of storage are determined by the specific system operating parameters 28 well as the characteristics of the materia. being transported. Tenks For live slurry storage a: the head and tail ends of a long-distance pipeline system, agtated tanks are typically utilized, The number and sizeto be usedare determined bycomparing the plant and pipeline availabilities in addition to nornal engineering ¢rteria such 1s allowable soil loadings and an economic analysis of available sizes. If he pipeline is to be overated in batch mode, for the first few years of operstion, because of low throughput requirements, the size of the tanks may be determined by the minimum batch length chosen for the system. Tais in turnis a function of pipeline length. - Agitated orage may also be required at intermediate pump stations. The pur- pose would be to facilitate pipeline section reconnection during restart operations as well as held slurry which has been flushed from stition piping during sump change-out. Ponds Ponds are utiized for the storage of high volumes of solids over intermediate to org periods. Ponds are generally 1sed when the storage requirerrent exceeds 24 h cf pipeline iow. Their prime advantage is that no eneigy is required to maiatain continuous stspension of the solid. Slurry storage ponds san be clasified into two basic types: semisctive and Jead storage, Semisctive ponds are equipped for recovery and reslurryingon short notice. Depending oa the design, it may te necessary to remove all the stored material beiore refilling the pond. ‘The major consideration in pond design is recovery of a uniform solids pasticle size distribution. Segregation of the solids can occur duiing the fillng or reccvery operation. The resulting coarse anc fine slurry slugs canbe very difficult to handle in ‘he downstream processes. Thiee major sccovery methods have been utilzed: mechanical, ckedge, and the Marccnafio technique. Mechanical recovery uses conventional earthmoving equipment to remove the settled solids. Prior to tke start of :ecovery operations, ‘he bulk of the free vaterSLURRY AND SLUDGE FPING c.607 must be removed from the pond by evaporation, ratural drainage, sand points, 6t an underdrain system, Dredge recovery requires the mairtenance o! a water lyer over the settled solids to float the dredge, The recovery dredge may be maintzincd on site or moved in for the recovery operation, depending on whether semiactive or dead. storage is required. ‘The Marconaffo technique” wes originally developed to remove setted slurries from the hold of ore sips. A high-presure waterjet on a rorating heac undercuts and reslurries the solids, and the material flows into an undercrain system for pumpirg away. ‘Slurry Pumps. Centrifsgal as well as positive displacement pumpsare available for pumping slurry. ‘The maximum pressure capability and maximum particle size limits fer different types of pumps are shown in Fig. C11..7. ‘The positive dsplacement pumps cin be diviced into piston, diaptragm, and plunger pumps. Fiston pumps can be used for slatively less abrasive materials while the diaphragm and plunger pumgs are used for handling abrasive slurries at high pressures. Tre initial capital costs and maintenance costs of positive displace- ‘ment pumps are higher than those of cen-rifugal punps, but their hydraulic efficiency is 85 percent, compared to about 60 to "0 percent ‘or the cen:rifugal slurry pumps. The flow rate per pump is limited in the case of positive displacement pumps while \ a T yt ~ = oa | (esnaanson vee ; i : 3 i § £10] Kaa 2 ‘cent fugel UY id 0 1c (50 Monumum sertile ize (am) FIGURE C1117 Maximum particle siz and pressum capabilities of different ypes of pumps.C608 AIPING SYSTEMS the head developed per pump is limited to about 30 m in the case of centiifugal shrry pumps Centrifugal Sturry Pumps. Centrifugal pumps are exensively wed for pumping sharry undex relatively low pressure, The main advattages of tiese pumys are as follows: 1. High flow rates can be achieved with a single unit ata reletively low initia cost. Pumps capable of pumping 2500 gpm (5680 m*/h) are available, 2. Very few moving and wearing parts are involved. : 3. They are simple to operate and maintain. 4, There is no practical estriction on the maximumsize cf solids that can be handled, 5, The flow through the pump is pulse-free. 6. They require little space. . 7. Valves art not required for the éperation of the’ pump. (Conitrifugal pumps have the following shortcomings: : 1. The maximum discharge head is limited t6 léss than 130'tt (40 m) for a éngle- “stage pump. With several pumps in series, a maximum’ pressure of about. 750 “psi (5.17 MPa) can be achieved.” : eee 2. The flow -ate of the pump is governed by the system pressure, 3. Attrition of friable solids may occur due to the high velocity of flaw thiough the pump Attrition can become important if the slurry has to pass throagh a ausaber of pimps. - - il 4. Seal liquid is required for good packing life. The seal liquid dilutes the surry. The amount of dilution could become significant whea the slurry passes though a number of pumps, ‘5. Centrifugal slurry pups are male robust because of tie abrasion of parts coming in contact with slurry, resulting in a low pump efficiency. ‘The centrfugal pump parts exposed io wear from slumy are the casing, impeller, * and gland seal. Mecharical shaft seals often are ineffeaive in slurry installa ions, A seal that incorporat:s a liquic fhish to keep solids from entering the gland isnecessary. “ To obtain good serv'ce, the casing and impeller should be linec with abrasion- resistant material. Both rubber-lined and “Ni-hard” (metal-iined’ units are used extensively. The size of the solids to be pumped determines the tipe of pump to beselected. Rubber-lined pumps are generally used withparticles u> to about 1.375 in (9.5 mm), and Ni-hard pumps ate used for coarser slurries. Howver, if ma‘erial wih sharp cutting edges is being pumped, such as crushed glass, Ni-hard pamps can be used 2ven for relatively fine solids. Rubber-lined pump parts usually have loager service life with fine materals, Coal is one matc:ial for whch Ni-haré line pumps have shown better parts life than rubber-lined tnits. To obtain good pump parts life,it is good practice to imit the impeller tip peed toless than 4000 fi/inin (1220 m/min). The pump parts ife on units running faster thin this speed drops in proporticn to the square of the impeller ‘ip-speed.SLURRY AND SLUDGE MIPING c.608 Reciprocating Pemps, Reciprocating pumps have several lesirable features: * The flow rate cf the pump is indepeadent of ssstem pressure. * The can meet any reasonable system discharge: pressure requirement. Pumps capadle of procucing 2300 psi (15.8 MPa) pressure have been used in magnetite pipelines. Units capable of discharge pressures greater thaa 5,000 psi (34.5 MPa) are wailable. * The overall efficiency of the pump, including drivetrain, s relatively high—on the crder of 85 percent, + Pipeline flow rete caa be determined without the use of a flowmeter. ‘The following disadvantages are associated wih this type of unit: * The maximum fow rate per pump is limited to less than about 3900. gpm. Further- more, this capacity is only available at relatively low discharge pressures. There~ fore, a large number of pumps operating in psrallel are aceded to handle the high Now rates and working pressures found in large longdistance systems. For example, seven 1,250-hp units are used at eaca pump station of the Samarco system. * Initia capital costs and maintenance costs are isually hign. Skilled labor is re- quired for operition and maintenance. + Variable-speed drives are needed to vary flow mates. * The fow throug’ the pumpis pulsating, which requires greater attentioa to station piping design to avoid vibration and ‘atigue prcblems. * The maximum size of particles that aii be pumped is restricted by the check valve seal requicements. - For material wth a maximum partick size of less than 0.Lin (2.5 mm) and with discharge pressures up to about 2000 psi(13.8 MPai, either piston or plunger pamps can be used. Slurry with a maximum purticle size of 0.1 to 025 in (2.5 to 6 mm) may also be handied with these types cf pumps if special design pump valves are used, The decision te use piston or plunger pumps is usually besed on the results of ‘@ Miller abrasivity test. Material with : Miller number below 30 can be handled using piston pumps, and material with a Miller number above 50 should te pumped with plunger units Between these values, the type of pump to use is bised upon other considerations: * Pistox pumps cen be of double-acting design sc that about twice the flow rate can be obtained for the same physica. pump size. * Plunger pumps are more adapatable toftshing and lubrication. A flushed stuffing box can prolong parts life. However, aflush fluid free of solids must be provided, and some diluticn of the slurry will result. Piston Pumps, F gure C11.18 shows the fluid end of a conventional pision pump. With such a unit, major wear occurs on the pistors, valves, cylinder liners, piston rods, and packing, To reduce maintenance, several variaiions of pision pumps rave been developed that limit the number of working parts in contact with the slurry. One such type is the diaphragm pump, the fluid end of waich is sketched in Fig C11.19. By pumping¢.610 PIPING SYSTEMS PACKING VALVE --LINER VALVE srurFine Box (OnCHARGE) BUSHING —_ A PSTONROD FLOW PISTON sueriew FIGURE C1118 Plunger pump fuid ene PULSATION DAMPENER > piscuarGe YaLve 40, DIAPHRAGM | SUCTION VALVE 9 F “HYDRAULIC FLUID, PISTON FIGURE C1119 Diaphragm pap fii end.~ SLURRY AND SLUDGE PING cent hhyéraulis fluid with a piston pump, the diaphragn is alterrately squeezed and expanded. With this type of pump, the valves and diaphragm are the only parts that experience major wear. One disadvartage s the possibility of diaphragm failure. However, this can te minimized through careful puarp design bs limiting the amount of diaphragm flexing andby selecting the membrane slastomer carefully. In addition, suppliers offer visual and audible monitoring systems to detect diaphragn failure. Units are available that will limit the amount of slurry entering the propelling fluid chamber. Plunger Pumps. Because abrasive slurries can greatly reduce the life of the pistons and cylirder liners of conventional pistor units, plurger pumpsare often sed. This type of sump maintains a clear liquid barrier between the pluager and packing by means 0! a flushing system, as shown in Fig. C11.2), Major wear is limited to the plunger, valves, and packing. VALVE PACKING. THROAT BUSHING. (DISCHARGE) PLUNGEF GLAND BUSHING A tanreannines—/ Lush tiauic FLOW ICTION) FIGURE C11.20 Plunger pump flu end. Hydroheists, The limitation of maximum particle size for positive displacement pumps aid the limitation of pump discharge presse for centrfugal pumps can be eliminated by using a hydrohoist system. Figure C11.21 shows a sketch of a hy- drohoist pumping system. The hydrohoiit consists f one or more chambers that can be filled with solids either in dry fom (Iock-hepper) or as a slurry. Once the chamberis filled with solids, water under high pressare is admated to the chamber to push the solids into the pipeline. A major application is foundin verticaltranspor- tation of coarse ore from deep mines. The advartages of these devicss are as follows: + Large solids can te pumped underhigh pressure without having them pas through a number of centrifugal pumps. The altrition of solids is thereby reduced, + A high-pressure water pump, having relatively igh efficiency compared to a centrifigal slurry pump, is used.ewiz PING SYSTEMS <= MIXTURE OUTLET — »)) CLEARWATER euner 4 HOH messunE TURBINE PUMP SLURRY MixiNG DEVICE FIGURE C12 Sketch oa hydrohois. * The number of moving pais exposed to slurry is consferably les than that with reciprocatiag units. * The life of slurry check valves is extended because of the reduced frequency of operation compared 1) reciproceting pumps; normally the valves open and slose in the presence of water. * A.larger capacity can be obtained with a hydrohoist than with a reciprocating pump operiting at the same discharge pressure. The disadvantages o° the hydrchoist system are as follows: * A sophisticated control system ii needed to open and close the slurry valves. ‘The flow rete of this device depends upon the system discharge pressure since a high-pressure centrifugal water pump is used. Pump Power The pump power is computed from the following 2quation: @ up= 2 F x effcienay (C1146) wlere HP = pumping power, KW (hp) Q-=flow rate, m'/h (gpm P = increase in pressure across the pump, Pa (2si) F = conversion factor, 1714 hp/(gpm-psi) {3.6 < 10° kW/jm'/(h-Pa)}SLURRY ANDSLUDGE PING 61a Example C113. Select the type, number, and power of pumps based on the fol- lowing: Flow rae, 3000 gpm (680.m'/h) Rise in head across pump(s) 260 ft (60 m) Specific gravity of slurry 1s Solution. ‘The pressure rise across the pumps is 260 ft X 1.5 x 62.4'144 psi = 169 psi (1.177 MPa). Select centrifugal slurry pumps Assume three pumpsoperating in series to get a total head rise of 260 +t (80m). Assuming a pump efficiency of 70 >ercent, tke power fer pump & obsined as follows: 3000 x 169/3 1714 x07 = 141° Use 150-hp motors. 680(1.17)/3 ~ 3607) = 106kW HP = Life of Pump Pars, Typical parts life for the different types of slurry pumps is given in Table Cil.5. The actual life depends upon the punp speed, discharge pressure, and abrasivity of the slurry. * TABLE C11.5 Pump Parts Life Partlife Pump part Piston —_-Plurger___—Tiaphragm Centrifugal Piston Rod 3000 ah ala ala Plunger sleeve ala 7m nia nla Piston 4000 ah ala nla Packing 6000 40 wa iso Diaphragm li nA 8000 nla Valve seat 1000 80 ‘5000 nia Valve body 1000 80 3000 n/a ‘Valve insert 1000 50) 3000 ala Inpeller ala nls nla ‘6000 Casing nla An nya 6000 AO ‘Selection of Valves ‘There are many manufacturers and typet of valves offered forlow-pressure slurry service, These designs are backed by years of experience in the minerals and dredgingindustries. Selection of valves for use in postive displacement slury pump stations introduces some new parameters besides those that must be considered in314 HIPING SYSTEMS selecting lovepressure valves. These include vibration of the piping and packng of sdiids into tke void spazes in the valves. Generally speaking, in high-pressure duty where a fall opening is not required, lupricaied phig valves rave been applied. Depending on the service, the wetted perts of there valves should be hardfaced to minimize abrasive wear during the opening or dosing cycle. Gate or ball valves have been used where a full 10und opening is required. Both types introduce the problem of removing solids fron the body cavity ster the valve has been operated in slurry. Agaia, it may be necessary tohardface ome wetted parts to extend the valve life. Slurry valves require regular lubrication and maintenance to remain opecable. ‘An important feature in selecting valvesis the ease of disassembly for mainterance. Some ball valves, for instance, have to be sent to the factory to reaew the seats or seals. Some naintenance problems may simply be a matter of soids finding their ‘way into the valve workngs, with tie solution simply being disassembly and cleanup ofthe valve. Some slurries (Such at limestone) may have some cementation proper- ties. Therefore, it is good practice to regularly lubricae and exercise valves that are otherwis: infrequently used. Abrasion Control Pipe abrasion is a major concern :n coarse solids pipelines. In temporary systems, the pipeline may'be replaced periodically. For example, the phosphate ninin, industry in Forids replaces tomporary steel pipes after bout 2 years of operaion. Use of special pipes, nonferrous pipes, and steel pipes lined with asrasion-resstant linings can significantly increase the life of pipes carrying coarse solids, Example C1L.12 addresses the subject of corrosion-erosion cont-ol for long-term service. Nonferrous Pipes. _Fotlow presstte applications, a nunber of non’errous materials can be considered for transporting coarse solids. Polyethylene. Ulteahigh-molecular-weight polyethylene has been extensively ‘used for tailings line construction. For sizes up to NPS 14 (DN 35)), itis available wth pressure ratings up to 250 psi 1725 kPa). This material isa maor improvement over unlined carbon steel or wood, and under good conditions it asts many times longer than :he materiel it replaces. Polybuty'ene.. Polybutylene ripe is a flexible thermoplastic pipe that has been developed relatively recently. Depending upon the diameter, its normally available with pressure ratings up to 200 psi(1380 kPa), although higher-pressure pipes have been made. Polybutylene has two major advantages compared to polyethylene: the tensile strength is higher, and it it less affected by extremes of temperature, Polyurethane, This type of pipe has been used for some in-plant systems, in particular, coal cleaning plants. The polyurethane is spun-cast into flanged spools up to NPS 12 (DN 300). Other Nonferrous Materials. Polyvinylehloride (PCV) polypropylene anc acry- lcnitrite-butadiene-styrene (ABS) have been made int> pipe but have poor wear resistance cmpared to polyethylene and polybutylene. They should not be consid- ered for lonz-life slurry applicaticns. Fiberglas: pipe is available with ceramic chips or te embedded into the inner surface. This material 's quite exxensive, and once the chips are broken ott, the epoxy matrir that holds the chips kas very poor wear resistance anc the pipe quickly fails. Also, jcint systéms designed to take highpressure a’e prone tofailure. Abrasive purticles attack the groove left at the joint, and early failure may result.SLURRY AND SLUDGE PPING cer Internally Lined Steel. Meny of the materials dscussed in the previous section can also be utilized as ¢ lining within a steel outer shell. By utilizing suet a system, the inherent pressure limitations of nonferrous materials can 2e overcone. In most cases, the use of a composite pipe leads to difficulty in joinng the pige sections Most systems require the use of stzel flanges with ‘he ining material sorving as the gesket material. In Situ Lining of Steel Pipe. Two methods of in situ internal lining of welded steel pipe have been developed. In one method 1 plastic pipe of slightly smaller diameter than the inside diemeter of steel pipe i pulled inide, and the annulat space tetween th: two pipes is grouted with cement. In another metkod a high- densitypolyethylene pipe of slightly larger outside ciameter than the insice diameter of the steel pipe is installed by pulling t after compressing it to reduce its outside diameter. The pipe expands and presies against the steel sipe upon release of pulling forces. ‘Theie liners are pulled inside the sieel ripe, vhich is welded prior to pulling the liner. The liner can be pulled to a distance of 1 km depending upen the pipe size and the number of bends in the steel pipe. In either mettod., best mechanical desiga and economic results arz obtained when lining is planned for in the beginning. In ar operating pipeline these lining methocs can be wed to extend service life. Instrunentation ‘The instrumentation used and,the major parameters measured and controlled are as follows: Pressure. Both tie suction and discharge pressures at a pump station are normally measured and cortrolled. Suction pressire is contiolled to prevent cavisation, The discharge pressure is controlled so as nol to exceed ihe maximum allowable pressure for the pump unit or the pipeline systsm. Bourdon-type pressure gages with an isolation diagram are generally used. Density. ‘The pumping head requirement varies with the chaages in slurry density. lithe surry density is not controlled within the pipeline design lim, the pipeline system can be overpressurized or the flow velocity zould drop below the deposition critical velocity. Slury density s normally measured by radiation gages. Density meters may be needed at each pump station in a multiple-station system. The density meters are normaly installed on the suction side cf a pump station. Density iscontrolled by adding dilution water toaigh-density slurry. Ifthe density is too low, the slurry can be retumed fo the preparation plént for concentrating. Flow Rate. Because slurries are often abrasive by nature, some imine flow- measurng devices such as orifice plates and pitct tubes are not applicable. For primary slurry flowmetering, magnetic Jowmeters are considered the best choice. Ultrasonic flowmeters sometimes are ased in applications including oil slurries where magnetic fowmeters are unsuitaole. Slurry Tank Leve! Measurement, A namber of diferent systems have been used for recording the level in slurry storage tanks.cwi6 PING SYSTEMS Conductance probes for level alarms and a pressure sensor mounted in the wall ofthe tank ate used to ebtain a rough indication of level of slurry ia the tank. This procedure has proved setisfactory although the accuracy of the level measurenent with a pressure sensor isaffected by changes in slurry dersity. A way to compensate. for this effec: is available but noi commonly used. Soaic devices provide cirect measurement of tank level. REFERENCES 1. Linssen, J.6.Th., “The Performance and the Future Development of Dredging Equip- ment,” Fint International Symposum on Dredging Techrology, University of Kent at Canterbury, England, BHRA Fluid Engineering, Paper AJ, 17-19th September, 1975. 2. Blatch, NS, Discassionof “Works or the Purification of the Water Supp of Washington, D.C,,""by Hazem, A. and Hardy ED., Trans, ASCE, vol. $7, pp. 400~), 1906. 3. Gregory,"Sumping Chy Slurry Tirough a Four-Ind' Pips," Mech. Engg. vol. 48,0. 6, 1927. 4. O'Brien, MP,, and Fokom, R.G., “The Transportation of Sand in Pipelnes,” University of Califérna, November 12, 1937. 5. Howard, GW., “Transsortation of Sand and Gravel'in'a Fourclich Ppe,” ASCE, vol. 104, 1939. 6. Caldwell, D¥., and Batbit, HLE, “he Flow of Muds, Sludgss and Suspersions in Cirular Pipes,” Trews. A. J. Ch E., vol. 37, pp. 237-66, February 1341, Be 7. Durand, R. “Basic Relationships o! the Transportation of Solids ia Pipé-—Experimental Research,” Proc: Interrationel Assee. for Hydraulic Research, Miaacapolis, Minn; 3spt. 1-4, 1953. i 8, Zandi, L, “Hydraulic Transport of Bulky Materials,” Paper |—-Advanees in Solid-L.quid Flow in Pipes and Its Application, edited by Inaj Zand, Pergaman Press, 1971: 9. Thomas, DG., “Transport Charactzristics of Suspension: VIII, A Note on the Viscosity ‘of Newtonian Suspensions of Uniform Spherical Particles," of Colioie Science, vd. 20, no. 3, pp. 257-277, March 1967. ° 10. Chong, J.S.,Christiansea, E.B,, and Baer, A.D., “Rheology of Concentrated Suspensions,” J. Applied Polymer Sei, vol. 15, pp. 2007-2021, 1971. Gay, E.D., Nelson, P.2., and Armstrong, W.P, “Flow, Properties of Suspensions: High Solids Concentration,” AICAE Journal, vol 15, x0. 6, pp. 815-822, November :969. 12, Hanks, R.W., ané Prat, D.R, “On the Flow of Bingham Plastic Slunes in Pipes and Between Parallel Plates” Sociery of Petroleum Engineers Journal, p. 342-346, Desem- ber 1967. fang Ricks, B.L., “Laninar-Turbulent Transiton in Flow of Pseudopistic eld Stress" J. Hydrorautics, vol. 8, no. 4, pp. 163-166, Cctober 1974 14, Dodge, D.W, and Metarer, A.B., “Turbulent Flow of Non-Kewtonian Systems,” AICE Journal, vol'5. ne. 2, pp. 189-208, fune 1959. 15, Stainsby, R, and Chilton, R.A, “Prediction of Pressure Losses in Turbulent Non-Nevton- ian Flows: Development and Application of a Hybrid Rhedogical Model,” Hydrorans- port, ol. 13.pp. 21-39, BAR A Fluid Engineering, Cranfield, Sedford, England, May 1996. 16. Hanks, RW, and Dada, BH, “Theoretical Analysis of the Turbulait Flow of Non- ‘Newtonian slurries in Hpes,” AICHE Journal, vol. 17, pp. $S4-557, Mey 1971. 17, Hanks, RM, “Low Remnolds Numser Turbuleat Pipeline Tlow of Pseuiohomogenzous Slurries," Hydrovanspat, vol, 5, pp. C2-23-C2-34, BHRA Fluid Engineering, Cranield, Bedford, Eagland, May 1978.SLURRY ANDSLUDGE PPING ce 16, Dodre, D.W., ant Meter, A.B., “Turbulent Flow 0: Non-Newtonlan systems,” AICRE Journal, vel. 5. 90.2, pp. 89-204, lune 1959. 19, Hanks, RW. ard Ricks, B.L., “Transitonal and Turbulent Pipeflow of Pssudoplastic Pluics." J. Hiydronautis, vol. 8, no. 1, pp. 18-23, Jasuary 1975; 20, Hants, R.W,, “Low Reynolds Number Turbulent, Pipeline Flow of Prendahomogeneous Sluries,” Hydrovansport, vol. 5, pp. C223-C2-34, BARA Fluid Engineering Cranfield, Bediord, Englanl, May 1978, 21, Tornee, B.Mck., South African Mechanical Engineer, vol. 13, p 89, 1963. 2. Wilson, KC. ard Thomas, A.D., “A New Analyss of the Turbulent Flow of Non- Newionian Fluics,” Can. J. Chem Eng.,vOl. 63. pp- 339-346, 1965. 23, Thomas, A.D. and Wikon, K.C. “New Analysis of Non Newtosian Turbulent Flow— ‘Yield Power Lav Fluids,” Can. J. Chem Eng., vol. 65, pp. 335-138, 1987. 24, Stainsby, R.,andChilton, R.A., “Prediction of Fressure Losses in Turbulent Nen-Newton- ian Flows: Devebpment and Application of a Hybrii Rheological Model,” Rydrotrans- ortvol.13, pp. 1-39, BHRA Fluid Engheerirg, Cranfield, Bedford, England May 1996. 25, Durand, R, “Basic Relationships of the Transportatien of Solids in Pipes—Emperimental Research,” Proc. International Assoc. for Hydraulic Research, Minneapolis, Ninn., Sept. 1-4, 1953. 26, Zandi, L, and Gevatos, G., "Heterogenecus Flow of Solids in Pipeline,” Proc. Hydraulics Divition, ASCE, vol. 93, pp. 145-159, May 1967 27, Wasp, EJ, et 2., “Deposition Velocities anc Spatal Distributon of Solids in Slurry Pipelines,” Hydrotransport, vol. 1, BARAFluid Enginvering, Cranield, Bedforl. England, September 1970. 28. Graf WH, Robinson, M., and Ucel, O.,"*The Criticel Deposit Velocity for Solid-Liquid Mixtares,” Hydretransport, yol.1, BRA Fluid Enginicring, Cranfcld, Bedford, England, September 1970. ‘ 29. Oroskat, ALR, ad Turian, RM,, “The Critical Vebsity in Pipdline Flow o: Slurtes,” AICKE Journal, ol. 26, n0. 4, pp. 550-558, July 198). 30. Wilson, K.C., and Judge, D.G.; “Analyticlly-Based Nomographic Charts for Sand-Water Flow,” Hydrotrauport, vol. 5, BERA “luid Enginiering, Crasfield, Bedford, United Kingiom, May 1978. 31. Wilson, K.C., “Deposition-Limit Nomogrims for Particles of Varicus Deasitesin Pipe Flow,” Hydrotrmspor, vol. 6, BHRA luid Engincering, Cranfield, Bedford, United Kinglom, September 1979. 32. Durand, R, “Batic Relationships of the Transportaticn of Solids in Pipes—Experimental Research," Proc. Intemational Assoc. for Hydraulic Research, Minneapolis, Minn., Sept. 1-4, 1953. : 33. Zandi, L, and Govatos, G.,“Heterogenecus Flow of Solids in Pipdine,” Proc. Hydraulics Divkion, ASCE, vol. 93, pp. 145-159, May 1967. 34, Turién, RM,, and Yuaa, T;, “Flow of Shuries in 3, pp 232-243, May 1977. 35, Newitt, DM, Richardson, IF, Abbott, M., and Turtle, R.B., “Hydraulic Cenveying of Solids in Horizontal Pipes.” Trans. Inst. Chem. Engn., vol. 33, p>. 93-110, 1955. 36, Babsock, H.A., ‘Heterogeneous Flow o/ Heterogentous Solids,” in Advances in Solid- Liquid Flow tn ripes and I Application, 1. Zandi, ed., pp. 125-148, Pergamon Press, New York, 1971. 37. Wasp, EJ, Regan, TJ, Withers, JG... Cook, PA.C..and Clancey, J.T, “Cress Country Pipeline Fiydraulcs.” Fipetine News, vol 35, pp. 20-28 July 1963, 38, Wasp, EJ.,Kennsy, .P, and Gandhi, R.L, “Solid Liqtid Flow Slury Pipeline Transporta- tion,” Trans. Teck Publications, 1977. lines,” AICAE Journal, vol. 23, n0.core PITING SYSTEMS 39, Lazarus, 1H, “Mixed Regime Slunies in Pipelines I: Mectanistic Moe,” J. Hydraulic Eng. vol. 115, no. 11, Arterican Society of Civil Engineers, p. 1496-1509 November 989, 40. Wilson, K.C. “Stationary Deposits and Sliding Beds in Pipes Transporting Solids,” #ydr0- transport, vol, 1, BRA Fluid Engineering, Cranfield, United Kingdon, Paper C3, pp. 28-40, 1970 Wilson, K.C. Streat, M.,and Bantin,R.A., “Slip Model Consetations of Dense Two Fase Flow,” Hydotransport, vol. 2, BHRA Paper bl, pp. 1-10, (972. Wilson, K.C, “Stationary Deposits end Sliding Beds in Pipes Trarspontig Solids,” Proc. Ist international Symposium on Dredging Technology, BHRA, Paper C3 pp. 28-40, 995, Wilson, K.D., “A Unified Physicaly-Based Analysis of Solid-Liquid Pipeline Flow,” Hydrotransport, vol. 4, Paper A2, pp. 1-16, 1916. Wilson, K.C, “Deposition Limit Nonograms for Particles of Varioas Deasitiesin Pipsline Flow,” Hydoiransport vol. 6, Paper Al, pp. 1-12, 1979, Wilson, K.C, “Evaluation of Interfacial Friction for Pipeline Transport Models,” Hydro- transport, val. 11, pp. 12-116, 1988 Wilson, K.C, Addie, GR., and Cift, R, Slury Transpo Using Ceatrifugal Punps, Elsevier Applied Science, New Yort, 199i, “ 41, Gillie, R.G,, Shook, C.A. and Wilsen, K.C., “At Improved Two Layer Model for Tori= zontal Slurry Pipeline Transport,” Can, J, of Chem. Eng., ol. 9, pp. 173-178, February 1991. : Jor a tes eimekey 42, Worster, Rand Denty, DF. “Hrdraulic Transport of Soid Metérial in Pipes," Proc: “Institute of Mechanical cngineerrig, vol. 169.1955. 2 43, Kao, D.T.Y., and Hwang, L.Y., “Critical Slope for Slurry ‘Pipeline Transporting oal and Other Solid Particles,” Hydrotransport, vol. 6, BHRA Fluid Engineering, Cranield, "Bedford, Exgland, September, 1979. "0 oh 8 “44 Turian, RAN, Hsu, FP "and Selin,S, “Friction Losis for Flow of Slur? io Papaine Bends, Finiags and Vales” Partctlate Science and Technology, vo. 1 n0, 4 Dp, 165~ SE Bee a and BOA D3 45. Gandhi, RL, Ricks, BL, and Ausle, TC, “Codirol of Corrotion:iosion in Sry Pipelines,” Ist International Conference on Inizrnal and External Proection of Fpes, “BHIRA Fiuli Engineering, Craniiek, Bedford, England, Paper G2, Seprember 1972 46, Derammelsere, RH. and Chapman,1P. 47. Linen; GP, “Marconilio—the Sytem and the Concept,” Bigineering anid Mining J., pp. 67-75, May 1970: he we : 48. Gandhi, RL, Snoek, PE,, and Camey, J.C., “Ai Evaluation of Slurry Puinps,” Proc. ‘th International Technical Conference on Slurry Transportation, Lake Taioe, Nev., March * 26-28, 1980 " : 49. Paddick, RR. and Stanan, 0.D,’ Pipeline: Transportation of Phosphate’ Slurries—A " Survey," Cdotado Schcol of Mines Research Institute, Mireral Industries Bulletin, vol. 20, no. 6, November 19,7, 50: Snoek, P-E. and Carney, J.C, “Pipeline Material Selection for Transport of Abrasive Tailings.” Froc. 6th Inernationat Technical Conference on Slurry Trassportation, Las Vegas, Nev, March 24-27, 1981.CHAPTER C12 _ WASTEWATER AND STORMWATER PIPING SYSTEMS . Ashok L. Lagvankar, P.E., DEE Vice President Total Wate: Managerrent Earth Tech, Inc. Oak Brook, IL John P. Velon, P.E. DEE Vice President Harza Engineering Company Chicago, IL This chapter’ prosonts information’ on the analysis and design of wastewater and stormwater piping systems. Wastewater piping sysems convey waterbome domes- tic, commercial, or industrial wastes 19 a point of discharge and/or treatment. Such systems are also known as sanitary sewer collection systems. Stormwater piping systems cenvey captured stormwater runoff to poinss of discharge. Such systems are also known as storm sewer systems. There are systems, chiely in older cities, that convey domestic, commerdal, and intustrial wastewater and storm- water munoff in a single piping system Such systems are celled combined sewer systems. ‘The design of both wastewater and stormwater collection systems mast comply with standards of city, county, and state as well as federal regulation agencies. Permits must be obtained for the disposal of domstic, commercial, and industrial ‘waste into receiving streams or water bodies, Stormwater discharges to receiving streams or water bodies may also requre permits ‘This chapter ‘ocuses on public ard/or private wastewater and stormwater collection systems generally serving a number of buildings. For information about in-building plumbing, which is net covered in this charter, see Cap. C13 of this book. The scope of this chapter does not peimit examiration of appropriate piping materials ‘or handling exotic wastes such as oils and acids. Refer to the index of this book for specific information on piping sysiems to ecnvey such wastes. 6.6196.620 PIPING SYSTEMS DEFINITIONS: A selection of terms used in wastewater and stormwater technology follows. Addi- tional terms are defined at the place of use Sewer: A pipe or conduit that carries wastewater or stormwater. Wastewater: The spent water of a community. From the standpoint of the source, it may be < combinaton of liquid and water-carried wastes fron resi- dences, commercial buildings industrial plants, and institutions, together with any groundwater, sarface water, and stormwater that may be present. The term wastewauer is often used instead of the less inclusive sewage. Stormwater: The portion of vater that runs off the ground during and inmedi- ately folowing a reinstorm, sxowmelt, or other flooding ever. Sanitary sewer: Asewer that carries liquid and water-cerried wastes from resi- . . d depth of flow : D : pipe diemeter fy f friction factor fos feet per second ft foot gped gallons 2er'capita 2er day apd gallons 2er day gpm gallons er minute AH hydraukc head hr hour in inch kg kilogram km kilometer * Ib pound Iped liters per capita per day Ipd . liters per day m meter m cubic meter med million gallons per day min minute mm millimetere822, PIPING SYSTEMS n dimensionless roughness coefficiert ao quantity of fow 7 hydraulic radius 38 slope of the energy gradient s seconds ov velocity WCF Water Pollution Control Federation (now callec Water Environment Fed- eration) QUANTITY OF FLOW URTV OP ROW 7 Quantity of Wastewater Flow. ) - Ore of the fist critical steps in the design of:sanitary sewers is thx’ éstimation of design flows of wastewater tributary to them. Sanitary sewers must be designed to Provide capacity for the present andestimated future quantities of demestic sewage, conmercial aad industrial wastewaters, and infltration/inflow. Dosign Perioé and Tribwary Flow.’ Lateral and branch sewers of ¢ sanitary sewer system should be designed for the ultimate population density to be expected in the area served. Main and trunk sewers are commonly designed to andle the flows to be expected from a population 25'to 50 years in the future anc in some «ases the ultimate buildup of the servicr area. The estimation of future flows can be arrived at only after a cetailed study of the land use, population growth trends, wa‘er consumption rates commercial and industrial growth, etc. Forecasts of popu. lation and the commercal, industrial; and institutional development expected to bem place at the end of the design period should be obtained from loval and regional planning orgmizations and adjusted when appropriate ty site-specific knowledge. Quantities of Flow—Overview. Wastewater tributary tc sanitary sewers emarates fren residental, commercial, and industrial land uses. The most inportant single. index in estimating flowsto the sanitary sewer from each of these land-use sources is tne rate of vater consumption, Fows entering the santary sewer generally can be estimated as the water consumption, less an ellowence for uses outside the building that typically do not return to the sewer (landscaping, car washing, etc.), plus an allowsnee for iniltration/icflow of clear water into the sewers. Note that ‘waler consumption means water delivered to the user. The quantity o' water acttally pumped into tie water distribution system includes water that will belost as leakage en route to wers. If the designer 1s pamping records available, as opposed to corsumption (or water billing) records to estimate water consumption, the pumping records should be adjusted downward by 5 percent to as much as 30 percent to account for leakage in the potable water system The quantities of waier consumption depend to some degree on the typ: of plunbing fixtures in residential and commercial buildings. Low-flow plumbing fix- tures, if prevelent in an area, will decrease water consimption aid consequent sewage flows. The presence of such fixtures should be accounted fer in estimetes. When estinates of sanitary sewage flow from small creas are being prepared,WASTEWATER AND STORMWATER PPING SYSTEMS ce22 the mostaccurate procedure is to make separate estimates of tie various classifica- tions of fow which make up the total. The classifications which are commonly used are domestic, commercial, and industrial sewage flows and iniltration/inflow. Quantity of Domestic Sewage. Domestic sewage flow, excluding allowances for infiltration from residential land uses, is generally equivalent tc water consumption minus aa accountiag for outside-the-hone water tse that does not flow back to the sanitary sewer (landscaping, car wasning, etc.). The amoust of water used for such outiide uses varies depending on climate of thearea and landscaping aractices. In temperate climotes, such a5 the midvest and esstern United States, ‘here is a distinct seasonal fluctuation of domestic water consumption dae to signifcant out- side water uses for landscaping, etc, Peakwater consumption occurs during summer seasons, and mininum consumption occurs during winter months whe outside vses are negligible. In such areas the wastewater fow rate, excluding infiltration, is best gauged by observing water consumption records during winter months. ‘In warmer, drier cimates where landsceping wate: use contiiues year-round, an estimated allowance-can be made during lowest-water-consunption moxths. The ratio of annual domestic sewage flow, ezcluding infiltration, t> annual water con- sumption (water billing, if available) ranges from 70 percent in arid areas with extensive landscaping water use to 95 percent in areas where litle orno landscaping, water us is practiced. ‘The average per capita domestic waer consumption varies from about 40 to 120 gpd (150 to 4501pd), depending uponthe characier of the area and the 2eonomic status of the populition. If water is supplied through meters, cecurate estimates of average per capite consumption can be made. If water is supplied uametered, estimates have to be based on the water consumpticn rates or measured sewer low ates which are known to prevail in other areas of similar character. Table C12:1 ‘presents typical westewater flow rates from residextial sources.” TABLE C12.1. Typical Wastewater Flow Rates fron Resdential Sources Fow, gal/(unt-d) Source Unit Range Typical ‘Aperiment Righ-tise Peson 35-75 30 Low-tise Person 50-80 65 Hoel Grest 30-55, 45 Individual residence ‘Typical home Peson 45-90 70 setter nome Feson 60-100 80 Linury home Peson 75-150 95 Cider home Peson 30-60 43 Simmer cottage Person 25-50 40 Moel ‘With klichen une 90-180 100 Without kiteben = Unit 75-150 95 ‘Traler park Peson 30-50 40 Voter gali(unit-d) x 3.78: = U(unit-d) Source:" Adapted in part fom Ref. 1.C624 PIPING SYSTEMS Quantity of Commercial Sewage. The quantity of sevage flow fom commercial areas varies widely depending upon the nature of the commercial activity. Allow- ares made for the quartity of sewage from commercial areas in large sewer districts are commonly in terms of gallons per day pet acre, gallons per day per square foot offloor space, or gallons per day per capita or per empbyee. Allowances mado by ‘wistewater Utilities for average wastewater flow from commercial areas Vary from 2000 to 60,000 gpd per acre (18 10 560 m°/d per hectare) and fron 15 to 500 gpd et capita (6('to 1900 Ipd per capita). Table C12.2 shows ypical average commercial TABLE C122 Average Commerdal Wastewater Flowt Avg. flow ‘Type of establishment (ed per capi) Stores, offies, and smi businesses 12-25 Hote 0-150 Motels 50-125 Drivesin theaters (three persons per car) 810 Schools, no showers, 8 period 3.35 _ Schools wit showers, & period 37-25 ‘Tourist and trailer camos 0-120 Recreationdl and summer camps 20725 as Nate: syd per cap Ind per eit, ‘Source: Adapted in par from Refi 2s ! flows for varous commercial categories? It is évident from the wide range. of estimates shown in Table C12.2 that for any area in which commercial activity, is animportant ‘actor, the estimate of sewage flow should be based ona special cudy of ‘he area. Quantity of Industrial Sewage. Tae flow of sewage from industrialestablishments may be purely sanitary sewage, or i: may also include waterborne intustrial wastes, Estimates of the sanitary sewage ac madc by the procedures alreidy considered, ‘The quantity of industrial wastes can be determined only by special studies of the individual industrial activities, ‘ Quantity of Infiltration. The rat: of infiltration of groundwater into sewe's is influenced by the size, age, and condition of the sewers; the position of the sewers with respect to the grourdwater table; the character of the soil; and the amourt of Precipitation. The infiltration rate for any one system will vary fom season to season. Extensive work has been done to quantify and renove excestive infiltration in existing systems in the United States since 1975, Therefore, sever monitoring results of inftration are often avilable for existing sewer systens and should be consulted. Specificaticns for instellation of xew sewers limit infltiation to between 100and 500 gpd/in-dianeter/mi (90 and 460 ipd/cm-diameter/kn) of sewer at the tim? of testing. Allowances for infiltration at the end of the design period ae higher than theinitial test allowances. Design allowances for infiltration for new ‘sanitary sewers at the end of the design period range from 3000 to 10,000 gpd/in-diameter/mi (2800 to $200 Ipd/em-diaraeterikm) of sever for a well-maintained sewer system and as high as 40,000 gpd/in-diameter/mi (37,000 Ipdicm-diame‘er/km).- | WASTEWATER AND STORMWATER MPING SYSTEMS c.625, Quantity of Inflow, Inflow is clear weter which enters the sanitary sewer from sources such as roof leaders, yard and area drains, foundationdrains, cocling water discharges, drains from springs and swampy areas, manhole, covers, cross-connec- tions be:ween storm and sanitary sewers,or surface water runcff. Inflow is generally associated with periods of rainfall, snowmelt, and surface water runoft. In many existing systems, inflow is the single largest flow component on rainfall days and is often responsible for the backup of wastewater into basements and homes, or the bypassing of ustreated wastewster to nearby steams and water courses. Inflow rates heve been found to vary greatly ‘rom systen to system and even within a given system owing to, the age and condition of the sewer mains and laterals and the type-of plumting practices which Lave been employed. Typical inflow rates experienced cange from 50 gpd/ft (600 ipd/m) of sewer length to over 1000 gpd/ ft (12,0(0 Ipd/m) of sewer length on a mzximum-hour basis. Insome systems, inflow results in total pesk flows 7-to 10 times the averege dry-weither flow. Owing to the high variability. of inflow and the d:amatic impact it.can have.on wastewater ‘conveysnce syster sizing, engineering studies are gonerally recommended to estab- lish the existing inlow conditions and the appropiate inflow allowance for future service extension. : = Quantity of Exfilation. Exfilization is the process by which wastewater inside the sewer flows outof the sewer through joints, pipecracks, et, into the surrounding bedding. Exfiltration is generally not a concern with respect to estimating wastewater flows because sanitary sewers generally do not fuxction under pressurized condi- tions. However, exfiltration is important to consider in conjurction with protection ‘of grouncwater in pressurized tunnels. Ia near'surfice open-channel sewers, signifi. cant exiltration cin occur only when the’ sewer is in. gross disrepair ia. granular soils or is. cross-comnected to a storm sewer or oer Outlet. In cases’ vhere flow ‘monitoring indicates exfiltration, investigation of the situations warranted to deter- mine the flow path of sewage. Flow Variations.’ Sewers should be designed’ to handle peik flow rates. that are expected to occur at the end of the design period. It is also desirable to.design sewers 30 as to minimize the problem of solids deposition during the early years of use, when the flows may be much lover than th» future laws. In mos cases the ‘two design objectives cannot be achievedsimultaneously (espetially whenthe sewers ‘are circular). Priosity of the design objzctive should be established on a case-by- case baiis. The flows vary from day to day and fron hour to hour withir each day. The rafio of the zbsolute maximum fuure flow mte to the initial minmum rate may vary from about 3 to 1 for large sewers serving highly developed areas to more than 20 to 1 for small sewers serving areas still under development. Figure C12.1 presents commonly used formulas for predicting ratios of maximum ‘o average flows and minimum to average flows as a function of connected population? These formulas assume cry-weather concition: (without excessive irflow) and n0 unusual industral use patterns. Where nordomestic flows nake up asignificant amount of flow, the flow patterns should be considered separately. Fixture-Unit Basis of Design. For very small tributary populations andtor institu- tions such as schools, hospitals, hotels, and factories, the required capacities of sanitary sewers may be estimated from the fixture-vnit flow rates defined by various local aad nationa. plumbing codes. The designer should consult local applicable plumbing codes for estimating such flows.628 PIPING SYSTEMS oa os. 05; oa 03 ‘Hatio of minimum oF peak ‘verage del sewage ove Twinimars tow 02| oul 232 486 7e9IC m9 8040 60 80 10 200 400 008001000 Fopuiaton, in thousands : Curve A shiree: Babbin, H.E,, "Sevarige and SewogeTrestment” 7th Ed, Ji Wiley & Sows, Ine. New York (153). = " Curve Az iouree: Babbit, H.E., and Beumann, EA. “Sewerage ard Sowape Trestment.” &th Ed, John Wiley & Sons, ine, New York (3988), ‘Curve B Sturee: Harmar. W.G. "Foncasting Sewage at Toledo tnder Ory Werther’ Conditing.” Eng. Newe-Rec. 80, 1233 (1918) . a " fC Curve ©'Seuree: Younguown Ohio, port - . Curve D stutee: Maryland State Densrtment af Health eurve prieered in 481K, 8f Aapled Hydkatles.” 2nd Ee, Mesrow-Hil Book Co. New Yerk (105 Curve & sayce: Giff, AM. "Eetimatne Variations in Domeriie Sovege Flow ‘and Srwerage, 62,178 (1348) Curve F sas: "Manud of Military Conttriction.” Coras ef Engineers. United States Arm, ‘washngtor,0.C. Curve & sturee: Faic, GM, and Geye, J.C. "Water Supsly and WasieWater Dienors" et Ee, John Wiley & Sor, ine, New York (i854) Curves fa,8, and G wee conttructed es follows: 2S Bis Bases, “ “wo. cove 8, ae sn ut pin nt HGURE C131. Ratio of Extreme Fows to Average Daily How Compilsd from Vailous Sourees. (Adaoted from Ref 2.) Requirement: of Regulatory Agencies. Some state regu atory agercies have e:tab- lished definite per capite flow ratesto be used when detailed studies and estimates of expected flows have not been rade. The designer should look into munizipal ani state regulations to obtain such regulatory requirerents. ‘Typical ofsuch regulétions in the United States are the recommended stancards of the Great Lakes—Upper Mississippi River Board of State Sani:ary Enginzers,WASTEWATER AND STORMWATER PPING SYSTEMS ce27 ‘These standards, uied widely by state agencies in the United States in regard to the design of sanita:y sewers, are as follovs.* The latest edition should be consulted. Design Period. In general, sewer systems should be designed for the estimated ultimate tributary population, except in considering parts of the systems that can be eadiy increased in capacity. Simila-ly, consideration should be givan to the maximum anticipated capacity of instituions, industrial parks etc. Design Factors. 1n determining the required capacities of sanitary sewers, the following factors should be considered: 1. Maximum hourly rate of sewage flow 2. Additional maximum rate of sewage or wastew:ter from industrial plants 3. Maximum rate-of groundwater infiltration Design Basis.. Two methods are commonly used: '* Per capita flow: New sewer systems shculd be designed on th: basis of aa average daily per edpita tow of sewage of nvi less than 100 gallon: per capiti per day (gped) [380 liters per capita per day (Ipcd)]. This igure is assumed to cover normal infltration, but a1 additional allowance should bemade whee ground conditions are unfavorable. Generally the sewers :hould be designed tocarry, when running fall, not less thanthe following daily per capita contributionsof sewage, exclusive of sevinge or other waste from industrial plants. * Lateradls and submain sewers: 400 gpec (1500 Ipel). + Main, nk, dnd dual sewers: 50 gped (950 Iped). +. Interceptors: tntercepting sewers, in the case Of combined sewer systems, should fulfill the’ above requirements for trunk sewers and have sufficient additional capacity to care fpr the necessary increment of stormwater. Normally nointercep- tor should be designed for less than 250 percent of the gauged or 2stimated ‘avetage dry-weaher flows, Fo.” ‘Alternate method: When deviations fron the foregsing per capita retes are demon- strated, a brief description of the procedure used for sewer design must be included ia the design report. ‘Summary of Considerations for Determination of Sanitary Sewer Capacity. The. following summary of the consideretionsnecessary for the dete-mination of sanitary sewer cepacity is given in the American Society of Civil Engineers (ASCE) Manual of Engireering Practice, No. 37 {Weter Pollution Control Federation (WPCF) Man- ual of Practice No 9], “Design and Corstruction of Sanitary and Storm Sewers.” ‘The quantity of wastewater which must be transported is besed on fullconsider- ation of the following: . 1 The design perod during which the predicted maximum flow will rot be ex- ‘ceeded, usually 25 to 50 years in the future. 2. Domestic sewage contributions basedon future population and future per capita ‘water consumption; or if amore satisfactory parimeter thaa water corsumption is available, tha: parameter should be used. Carzful analyss should bz made of poptiation distributions and the rebtionship of maximum and misimum to avercge per capita sewage flows. The fixture-nit method of estimating peak ratesshould be employed for small pooulations, giving due care to the forecasting of th: probable number of fixture writs and weter use per capita. When largeces PING SYSTEMS areas are tc be considered, the peak rate of flow per capita or per acre sometimes 's decreased as area and populaion increase. 3. ‘n some instances, maximum flow rates may be determined almost entirely by extraneous flows, the source of which may he foundation, basement, rooi, or sreaway drains; storm runoff entering through manhole covers; or infiltration, Foundatior, roof, an¢ areaway drain connections to tanitary severs should be prohibited, Proper corstruction and yard-grading practices shouldbe mandatory. Nevertheless, there may be times when strict prohibition may no: be feasible or gven practicable, In any event, some stonmwater ard surface water will get into separate sanitary sewers, ad a judgment allowance, therefore, must be nade. 4, Commercial area conzibutions are sometimes assumed to be adequately pro- Tided for inthe peak allowance for percapita sewage flows in smal communtiies. A per acre allowance for comparable commercial arecs based on records is the ost reasonable approach for larger communities, 5. Industrial waste lows should incude the estimated employee contribution, esti- nated or gaged allowances per acre for industry as a whole, and estimated or tctual flow rates from planis with process wastes which may be permitted to enter the senitary sewer. 6, Institutional wastewaters are usually domestic in nature although ome industrial vastewatermay be generated by manufacturing operations at prisons, rehabilita- tion centers, etc. . 7, air-conditioning and industrial cooling waters, if permitted to enter severs, ‘nay amount to 1.5 to 2.0 gpm (C1 to 0.13 V/s) per ton of non-water-conserving cooling unis. Unpolhted cooling water should be kept out of separate sini- tary sewers, 8. Infiltration may occur through defective pipe, pipe joints, and sructures. The rrobable anount should be evaluated carefully. Design allowarces should be larger (under some circumstances very much larger) than those stipulatec in ‘constructior specificatons for which acceptance tests are perforned very soon ater construction. Uncerevaluation ofinfiltration is one reason why some sewers have become overloaded, 9. ‘The relative emphasis given to ezch of the foregoing fectors varies among exgi- neers. Some have set up single Values of peak design dow rates for the varinus Cassifications of tributary area, thereby integrating allcontributcry items. It is recommended, however, that maximum and minimum feak fows ised for design purposes be developed step by step, giving appropriate consideration to eich ctor whict may influence design, Exanple C12.!; Design Example for Calculating Wastenater Flows Determine the average and peak was:ewater design flows for the dowastream eni of a sanitary sewer designed to serve an area of 340 acres. Current resicential poplation is 1000 (primarily single-family homes), Proected ultimate populstion in the future is 3500 (G00 people ia single-fanily homes and 500 people in arartmenis), Additionally, conmercial and institutional employment in the service ares cusrently ¢150empl>y- ces, primarily in office and non-Water-intensive businesses Future commercial end institutional employment in the servce area is projected to be $00, again in similar types of businesses. There is negligible existing industrial development, but an 30- acre industrial 2ark is planned. Future number of employees working in the indas- ‘ial park is estimated to >¢ 800.WASTEWATER AND STORMWATER PPING SYSTEMS C4629 Solution. ‘The solution proceeds stepwise throagh the parameters to be con- sidered: 1. Selectthe design period of the sewer. Fer this sewer, because the area and popula- tion is small, the design condition is the ultimate future coxdition. 2. Estimate the average unit rate of flow of domestic sewage from residensial areas. Based on water billing records of evisting resitential single-family accounts, average water consumption duriag winter montls (November to January) is 80 gped ind during summer months (Jure to August) is 110 gpcd. ‘Apartments in adjacent communities have vinter water unit consumption rates of 65 gpd. Average wastewater fow from the residential sector istherefore 80 gped (assurned to be same as water consumption during vinter monchs) times 1000 2cople, or 80,000 gpd for existing conditions. For future conditions, the averaze flow is 3000 people times 80 sped for single-family residentia, plus 500 people times 65 gped for apartments, which totals to 272,500 gpd. =~; : 3: Estimate Commercial and institutional vasiewater flows: By using Hiow rates from. -similar ‘establishments in adjacent conmunities it way found that an.average per employee flow rate of 25 gpd was appropriats. Current flow from this sector is estimated to’ be 3800- gpd, and future average flow would be 900 employees ‘ctiimes 25 gped, cr 22,500 gpd. oe % ue 4, Estimate the industrial wastewater flows. Curreat industrial flow is xegligible. Projestion of fuure flow requires a thorough analysis of tte plaaned industrial park. It wes detzrmined in consultation with the developer, muni officials, and the zoning board that the ultimete future industrial rark would have 800 employees and one significant water intensive irdustry. The water-intensive in- duistr) ‘will discaarge: process water fo’ thé sever. at. the rates of 4,000, gpd (average) aiid 80,000 gpd (peak hour} Total average future flow is estiiatéd to ‘be 8M) employers times an allowance of 25 ppd rer employee (20,000 gpd) plus 40,00) gpd for process water. These two factérs total 60000 ‘gpd. cf average ‘tne Bowen ee eee rs toa OO weg ax 5. Estimate an allowance for the peak cuantity: of future infitration. Because. ‘sewe:s are of new construction, the future allowaice for infitration willbe 15,000 gpd/ni of sewe:. Estimate sewer length roughly at 1 mi ct sewer per 25 acres ‘of service area. Estimated sewer lengtt for the 34-acre areais 13.6 mi. Allowance for infiltration will be 15,000 gpd/mi times 13.6 mi of sewer, or 204,000 gpd. _-6 Determine the peaking factor. The peaking factor is taken from Fig. C12.1 using ‘curve G. Where significant nonresidential sources are. present, it is common to substitute population equivalent for populaton in-the: equation. Fopulation equiralents of commercial and industrial flow ave estimated by dividing the per ‘employee wate: rate by the per resident water rate and multiplying by the number of employees. For the future condition, the popalation equivalent of the commercial industrial component is 23 gpcc divided by 78 gped (weighted- average rate of single-family and apartment we) times ‘700 empleyees (900 cominercial anc 800 industrial), or 5¢5. The total population equivalents in the future will be 3500 population plus 545 population equivalents for commercial and industrial sector, or 4045. The peaking factor from cu-ve G woud be 3.33. 1. Estimate peak design flows. Peak design flows are calculated by first aseessing the averige flow rate from residential, ccmmercial, and industial sectors excluding infiltration. This would be 272,500 gpe (residential) plus 22,500 gpd (commercial) plus 20,000 gpd (industrial), for a totel wastewater flow of 315,000 gpd. MultiplyC630 NPING SYSTEMS this sum by the peaking factor of 3.33 to obtain a peak hourly Jow of 1,040,000 gpd. Add to this the peak allowance for infiltration of 204,000 gpd and peak flow of process wate: from the water-intensive industry of 80,0(0 gpd. The total peak hou: design flow would be 1,334,000 gpd. QUANTITY OF STORMWATER FLOWS Stormwater runoff is the portion of precipitation wheh flows ver the giound surface during and after a precipitation event. The design of storm sewer: and combined severs sequizes the estimation of design stomnwater ficws. There are a number of methods to develop peac stormwater flows. Among thest are the rational formula, the hydrograph method, the inlet method, and tke unit hydiograph method. ‘These basic methodologies have teen incorporated int» a number of proprétary and public-domain computer programs. Among the public-domaia programs, Ili- ncis Urban Drainage Area Simulator (ILLUDAS).and Stornwatzr Management Model (SWNM) ere in common tse in the profession in the United States. Discussion of each of the methads is beyond the scope of this chaptor. A3 such, discussion here is limited to 2 desaiiption of the rationa method, the simples and most basic analytic procedure in common use. Design Flows . The development of design flows for storm sewers is based on te area-specific poitical, ecoxomic, regulatory, and meteorological contitions. Typical practice is to design the system to accommodate the peak flows ofa rainfall event that bas a probability ofrecurzing every 3 to 19 years. For commercal and higl-value dist-cts, the design recurrence in‘erval could be as high as 30 to £0 years. The duration and intensity of that reinfall, establish:d from the area’s meteorological history are used to estimate the peck design flows, ‘The Rational Method Th: rational method is the most base procedure in common use for tie computztion of rates of stormwater runoff for storm sewer design. The rate of runoff Q in albic feet per secord (m/hr) s given by the equation Q=cia ~ (C2.1) Q=1CiA — SIunis (C12.1M) in which A is the size cf the draisage area in acres (hectares), iis the average intensity of rainfall during the design event in inches per hour (mm/h) for a duration equal to the time of concentration of the area, and C is 1 runoff coefficient whose value depend: principally upon the character of the are The assumption behird the raticnal method is that th: runoff rate at a point in a drainage basin for a given rainfal! intensity will increase and reaca its maxinum when the durction of the rainfall is equal to the time of zoncentraton of the area (the time required for the runoff to fow from the remotes: point of the contribuoryWASTEWATER AND STORMWATER PIPING SYSTEMS ceat area to the poisit where Q is being meavured). By this assumotion, the maximum runoff rate Q which can be expected to occur with any given frequency will be produced by a stom having the average rainfall intensity correspondiag to the given frequoncy and duration which equa the time o' concentration. The application of the rational method requires knowledge of the rainfall intensity-duration-fre~ ‘quency characteristics for the locality. A though the assumpticns are notstrictly in accord vith the mechanics of the ninofl pracess, tre rational method has proved to be a practical procedure for storm sewer design because the accumulated experi- ence has resulted in practicable values fer the runoff coefficiert. Reported practice generally limits the use of the rational methed to srban areas less than 5 mi (15 kn’), Fer larger areas, storm sewers designed using the rational method will result in an overdesign o! the storni sewer sizcs. Therefo:e, for larger areas, application ‘of hydrograph methods is usually warranted toavoidoverdesign of the stom sewers. Time of Concentration . ‘The time of concentration is the time required for the runoff to low from the remotes: point of tie drainage area to the poiat under design. This is the minimum, time nevessary to permit the drainage area to contibute to the flow anc generate the maxmum flow rate at the point. The time of concentration consists of the inlet time, ortime required for the runoff at the upper ead of the drainage area to reach the nearest inlet to a sewer, plus the time of flow in the sewer from this inlet to the point being coasidered. Jnlettime will vary with the nature ofthe surface (c.g., grassy, paved), the slopes, the nature. ofthe established drainage.channels such. as’ street gutters, and: the ‘antecedent conditions. Because the inlet time:is small, it is commonly:ehosen ‘on the basis of experience. In densely developed areas with a high percentage of paved surfaces and closely spaced inlets, an inltt'time as owas S'min may be considéred reasonasle. In'moderately developed urban areas with flat:Sopes, the inlet time ey vary from 10 to 20 min: In flat residential areaswith arelaively low petcentage of paved surface, ihe inlet time may be as high’ a: 30 min. It is possible to’ make estimates of the irlet time by calculatirg the time of flow over the varous types of surfaxes, but such estimates can rarely: be made with-a high degree-of accuracy. ‘The :ime of flo in the sewer is computed from the hydraalic properties of the ~ sewer, tie common practice being to use the average full-flow velocity forthe sewer computed for the prevailing'ground sicpe. Altermatively a velocity of 3 to 5 ft/s (1 to 15 m/s) can be assumed for initia calculations. The use of full-flow velocity in the calculation of the time of flow inthe seweris a conveaient and reasonable method of making the initial determination of tme of concentration for hand calculations. The cesign can de checked to account for variance in trave. times for partly full or surcharged sewers and rainfall pattems during the design storm. ‘The calculation of the time of conccatration is illustrated by exampl> C12.2. Exsampie C12.2. Determine the time of concentration for the drainage system shown in Fig, C122. Assume Manning’s roughness coefficient for the pipe sections to be 0013. Solution, The following information is derived from Fig C122: Pipe section AB: Length, 1000 ft; diemeter of tne pipe, 15 in (1.25 f); slope of pipes, 0.003; Mannings roughness cvefficient n, 0.013.C032 HPING SYSTEMS 1009 4. 1s'6 5 0.38 ‘e = 2.8 mutes FIGURE C122. Caleulatioy of time of concentration. Pipe sectian BC: Length, 1000 f: diameter ofthe pips 24 in 2; slope ofpipe , 0.002; Mannings roughness coefficient n, 0.013. - First, estinate the time for conveyance of runoff from the remotest catchnent area.to the remotest inlet to the sewer. In a moderately Jeveloped urban are, use 15min for flow from the: catchment area to travel to point A. Nest calculate the time for flowto travel from A to £, using Manning's equation, (C122). For metric units, Eq. (C12.2M) must be used. For pipe section AB, the parameters n ind s are 0.013 and0.003, respectively. Parameter r (hydraulic dius) mut be calculated. For pipe running, the full value ofr is mD*/4/r- D or D/4 where Dis the diameter of the pipe in feet (for SI, D must be expressed in meters). By substituting the vaues of the parameters in Eq. (C122), the velocity in the A-to-B section at full flow is 2.9 fp. The time to travel irom A to B is 1000 fi divided by 2.9 ft/s, cr 5.7 min, Similarly, the time to travel ‘rom B to C at a fullflow velocty of 3.2 f/s is ‘5.2 min. The time of concentration, therefore, to point A is 15 min, to point B is 207 min, andto point Cis 25.9 mir. For drainage system: describedin metric units, ust Manning's equation (C12.2M), Reinfall Frecuency It is usually prohibitive, on the basis of cost, to construct storm sewers capatle of handling the largest conceivable sorm. Current practice is to use storm raiafalls having an average expected frequency of once every 3 to 10 year: for the design of storm sewars in residential areas and storms of 10 to 30 years for commercial and high-valve districts.WASTEWATER AND STOIMWATER IIPING SYSTEMS 6.633 Rainfall Intensity Duretion-Frequency Relationships ‘The rainfall characteristics which must 2¢ known for storms:wer design are pre- sented in a concise manner by the intersity-curatin-frequercy curves, which can be prepared from a long record of predpitation et a given station, Figure C123 shows such @ set of curves prepared br the Illincls State Water Survey' for the northeastern area of Illinois for the years 1901 to 1983‘. This figure shows, e.g... that fora system to be designed with a tine of conczntration of 30 min ard a design ‘recurreace interval of 5 years, the total arecipitation is 1.3 in This meaas that the total rainfall of 1.3 in during a 30-min period will be equaled or exceeded in the area forwhich the curves apply, on averaze, once every 5 years. The rainfall intensity term in the rational method equation (Eq. (C12.1)] is expresied in unit: of inches per hotr. For the 30-min-duration, 5-year recurrent storm, the rainfall intensity would te 1.3 in divided by 0 h, or 2.€ in/h, Similar data tc those shown in Fig. C123 for many -cther localities have teen compiled and published. by theU.S. Weather Bureau. State agencies in cha:ge of stornwater management should. be consulted to obtain rainfall intensity-duration-frecuency. dats accepted for use in local arzas, as sued data vary widely by geographical region. oat 29 4 r nontHeast 1of- |} E ps 5 a Bob 3? J g | pn ca | ss" ] al | 2 on ors se 20 co 10 RECURRENCE INTERVAL (Years) FIGURE C123. Frequency distributions of rainfall for Norheast Illinois climatic secton forstorm periods of $ min to 10 days and recurrence intervals of 2 nonths to 100 years. (Fron F. A. Huff and JR Angel, “Fraquency Distrbutions and Hydroclimate Characteritics of Heavy Rainstorms in Illinos,” Ilinots Sate Water Survey, 1989.)casa pinino systems Runoff Coefficient Tre runoff coefficient 7 in the rational method formula ic the variable which is least susceptble to precise determnation. Whereas the ase of the runoff coeffcient plies that there is a constant rato of runoff to rairfall, the actual ratio for a given area will depend pon the condition of the area at the time >f occurrerce of the storm and will increase with the duration of the storm. A more logical procedure than the rational methed for storm sewer design woulc be to subract the rainfall losses due to infiltration and retention in surface depreisions and ‘0 distribute the remainder a: an actual hydrograph of runoff. Computer programs attempt 10 ac- ceunt for such parameters. However, because of the grect variability in the time distribution of the rainfal] itself aswell as the difficulty n estimating the quartities of infiltration and surface depression storage, the us: of the rational fomula sinply involves an estimate of the value of the runoff coefiicien: C. A common Pmctice is touse average coefficients for various types af districts, the coeffidents being assumed to be constant throughout. the storm duration ‘The: range of values for the runoff coefficient reported to be in common.we is shown in Teble C123. ' TABLE C123 Runoff Coefficients by Lind Use . o> Type of area ‘Runoff coeificient: C= =" o> Business: _ Downtown areas 90-895. Neighborhood cress 0508.70 Residential. i “Single-family'arcas 9.30.0.50 ‘Moliunits, detached: 0.40-0.60 Milliunits, atuached'= 0.60-0.75 Residential (subuiban): 025-140 . Apartnent dwelling areas 050-470 Industial Light areas 0:50-1:80 Heavy areas 0.80-090-. ‘ Parks, cemeteries 0.10-¢25 : Playgrounds <0 0.20-€35 b Railroads, yard areas 0.20-¢40" Unimgroved areas 0.40-€30 ‘Source: Adapted from Ref. 2, For specific small areas it is more logical to relate the value of C to the actual type of surface. Coefficients commonly used are shown in Table C12.4. When an area is made up of different types of surfaces, a common procecure is to se a weighted-average coefficient. ‘The coefficients in Tables C123 and C124 are designed for use for storms of 5-10 10-yr frequencies, For less-frequent, higher-intensty storms, the coefficents should be higher, because the infltration and surface retertion will be smaller Proportions of the total srecipitation. Licewise, for mofe-frequent, lower-intexsity storms, the coefficients should be lower than indicated in the tablesWASTEWATER AND STORMWATER PING SYSTEMS 635 TABLE C124 Runcéf Coefficients by Surface Characteristis Surface Runoff coefficient Streets Asphaltic 07-095 Concrete 08-095 Brick 07-085 Drives and walks 025-085 ~ Roofs, 075-095 Lawns, sendy soil: Flat, 2% 0.05-0.10 _ Average, 2 to 74% 010-015 Stecb, 7% ‘0:15-0:20 : Lawns, heavy sol: . et Flat, 2% ° 03-047 . : ’ Average, 2 to7%> " —0:18-0.22" g » + Steep, 1% 0:25-0:35!" ‘Note: Perceatagos ven fr lawn are average {ground sitace slope in he catcamant iran, Runoft ‘Coefficients for sects, crves, and roots are litle ‘affected bythe slope of be suracee, Source. Adapted fom Ret. 2 HYDRAULICS OF: SEWERS: ‘Where possible, itis common practice 13 design sewers to flew with a fee water surface. Hydraulically, this condition is termed opes-channel fow. The advantages of open-channel flew are twofold: (1) The free water surface vill allow ventilation of the sewers, and (2) the velocities at lewer flows zan be kept reasonaby high to * facilitate self-cleansing of the sewers. In areas where the depth of excavation for ‘a sewer ecomes ureconomically large, ¢ lift (pumping) station and force main are commorly installec to convey wastewate: to a locaton where gravity open-channel flow can resume. The force main is termed such becaiise sevage is “forced” by pumping and the flow is pressurized flowas opposec to open-channel(gravity) ow. Itis common practice to design sanitary sewers with slopessufficient t> provide for velocities of 2 £/s (0.6 m/s) when flowing full. Experience shows that with such slopes, trouble from deposits is seldom encountered. Storm sewers are commonly designec for a minimum full-flow velocty of 3 ft/s (1 m/s) in order to resuspend sediment deposited from intermittent storm events Although the flow in sewers is seldon steady o7 uniform, it is impracticable in ‘mest cases to take this into account, and each section of the sewer is usually designed with the assumption that the flow is steady and uniform. In certain conditions it is important to check the impacts of nonuniform flov. These ccnditions include the drawdown, backwater, and hydraulic jump conditions. These specialized conditions arc briefly discuss-d lator in this chaper. Specidized computer progiams that dynamically route flows can analyze sich conditions. Such programs, currently proprietary, shoulc be entering the pubfc domain in coming years.c.636 PIPING SYSTEMS. \ UNFORM-FLOW FORMULAS AND CALCULATIONS: ‘The most widely used formula for cabulating open-channel uniform fow in sewers is the Manning equation: 86 ” velsrsia tunis) (C1220 PBS (C122) where V = mean velocity, {t/s (m/s);n = dimensionless roughness ccefficient; r= hydraulic radius, ft (m), which is the wetted cross-sectional area divided by te wetted perimeter, and S is the slope of the energy gradiert, ‘The roughness coefficient n varies from 0.010 for smooh surfaces to 2s high as 0.10 ‘or-rough ratural channels. It is common practice to se velues of Manning's ‘nof(.013 for sever design. This value makes some allowance for the future condition of the pipe as well as for disturbanecs in the flow resulting from rough joints and interior coatings of grease or other metter. Figure C12.4is a diagram for the solution of the Manning 2quation applied to circular pipes flowing full, with n equal to 0.013. Pipes Flowing Full ‘The Manning cquation car be transformed to conveniently determine the quantity of flow of a cirailar pipe running ful at'a particular slope ~ . eet . 2 2. SAS pre “f ms (e123) .2= 0312 pen gin ‘SLunits (ci2a% n wher Q is the quantity flow at fill ow, (ais); Dis damets of pipe, (m); nd n and’ are defined’as above. Sinilarly, other converient wansfrmations of the full-tow pipe Manning fér- mula are as follows: ve 25% pissin (C124) v 22 Ds? (Stunits) (clz4m) se 46 1? (C125) ‘Ss ao ng? . (Stunits) (C12.5M)WASTEWATER AND STORMWATER PING SYSTEMS 6.637 106.9) 0! aco 2«0] 20 ico n gallons per day bé “s cal 03] «2 al 001 400 01003 ous “O10 0.02 00) 005 0.10 Slope FIGURE C124. Discharge of circular pipss (running fill) based or Manting formula! (rr whee Q = flow nte, A = cross-sectional aea, r = hydtaulic radius,$ = slope of energy gradent, and n = cocficient of roughnes = 0013, Curve @—Ninimum Sloves per “Recommended Standards for Sewage Werks,” Great Lakes—Upper Misissipri Ri Boaid of Stato Staitary Enginoers. (Adapud from Ref 3) QC438 PPING SYSTEMS pa 238 3183) = pie 18 3/8 (12.6) (Sl units) (C126M) where V is velocity of flow, ft/s (m/s); and D, S, Q, andi are defined as above. 18 Flowing Partly Full ‘Manning's equation and Fig. C124 are convenient for detormining flows and veloci- ties for pipes :unning full In determining velocities and depths of fow when pipes are flowing urder partialy full concitions, diagrams like Fig. C12.5 sre used, Figure CLS gives tke hyéraulie elements of ciscular pipes flowing partly ‘ull. This figure shows the ratios of the values of the various elements to “he values ‘or the flowing- ful condition. The cross-sectional arca and the hydraulic ridius are purely geometric furctions and hence. are independent of n. The velociy and discharge for any Particular ratio of deptt to diameter depend upon whether n is assumed to be constant or veriable with the depth. Velocity and dischaige curves computed trom both assumptions are shown.” a2: 34 36 y \ \ i 8 i & 4 3 i tale radive . é 1 0 0% 02 08 04 08 06 07 08 08 40 47 42 13 Hydraulic elements <2 Your a Ata Phat FIGURE C125 Hydraulic ements of circular sewers. (Adapted fiom Ref. &)WASTEWATER AND STORMWATEE PIPING SYSTEMS C639 ‘The use of Manning's equation, Fig. C12.4, and Fig. C12.5 is illustrated by example C123, Example C12.3. A 36-in sewer pipe (1i = 0.013) is laid on a slope of 2.00 ft per 1000 fi. Find the discharge capacity ard velocity of flow of ‘he sewer flowing full, and fird the depth of flow and the velocity when the flow rete is 20.0 cfs. Assume that rvaries wit the depth of flow. Solution, Using the transformed Manning equation for full flow pisin the full-flow discharge capacity is 29.8 cfs (r' = 0913, D'= 20 ft, § = 0.002). This checks against Fiz. C12.4 which yields a reading of 20.0 mgd (1.0 med ~ 155 cfs) Velociy at full fow from Mannings’: equation is 4,22 fis, as. calculated from Eq. (C124), : At20 cfs the sower is partially full. Referring to Fig, C12.5, we see that g/O ful, = 20,0'29.8 = 0.67. The term q isthe flow in cubic feet per second for fhe partially full condition being analyzed. Start at the bottom axis at q/Q = 0.67, and read uip to a pcint on the solid-line discharge curye (n, f variable wita depth). Refer 10 the left axé to cetermine that the ratio of -he actual depth to the full deptt (a/ full is 0.67. Then, reading across to the veocity curve, we see tat the mean velocity at d/D = 0.57 is 0.94 of the full-tow velocity. Consequently, for a ov of 20.¢f5, the depth of fow will be 0.67 times 36:n, or 24 in, and the mean velocity will be 0.94 times 422 ft's, or 3.97 ft/s. Fignre C12.5 aso shows the relative velocities required to obtain equal cleansing ‘of the pipe at all depths of flow. This is besed cn T. R. Camp’s analyses of the movement of granular materials ia open channels? The diag'am indicaies that if « sewer 1as self-cleansing- velocities under flowing-ull conditons; the. velocity; wil also be self-cleansing for-all flow conditions at cepths. greater than. one-half, the diameter: This concept is important for designing sewers that initially willxperience relatively small flows, but are installec to accommodate: laige flow in the future when fopulation growth occurs, In the zarly years the sewer. nay:never experience fullsflow conditions. The use of the velocity for the equal cbansing curve on Fig. C125 is best illustrated by example. Example C12.4. A 3%-in-diameter sewer is being installed to accommodate a future flow of 149 cfs. Design flow in the first 1C years of installation will be 2.5 ‘fs. The target full flow velocity for self-cleansing is2.0 ft/s. Determine the necessary velocity for equivalent scouring at 2.5 efs flow. “ Soletion. First, determine d/D for 2.5 cfs. This is dore by first calculating @/Q (25 cfs/14.0 fs = 0.18). Then real up from ‘he bottom axis to the discharge curve, and then read d/D from the lefi axis as 0.33, Next read to the right to the velocity for an equal cleansing curve 10 determite the Uspy/Via tatio. The term smu is that velocity required at a given depth (d) that wil provide equivalent scouring to that provided at full depth (d/D = 1. The Uso! Vea tatio for d/D = 033 is 0.74. If the target self-cleansing velocity at full flow is 20 ft/s, then an equivaknt cleansing velocity of d/D = (33 is0.74 < 2.0 ft/s, or 1.48 ft/s. Determine what spe is needed to provide 1.48 fv's velocity at d/D = 0.33. Velocity at d/L = 0.33 Is determinedby reading from the left axis to tie velocity curve aid then dovn to determine v/ Vis. The v/Vay ratio is 0.64. If the sel-cleansing velocity of full flow is targeted at 2.0 ft’, the actual velocity at d/D = 033 would bbe 2.0 x 0.64 ~ 128 ft/s. Because the ictual velocity at d/D = 0.33 of 1.28 {t/s isc.610 PPING SYSTEMS less than the tequired self-cleansing velocity of 1.48 ft/s, there is z probability of deposition of stids in the first 10 years. To remedy this situation, it will be necessary to teepen the slope of the sewer to be installed to provide a ful-flow velocity’ greater than 20 {Us Hovever, other ramifications of sleepening the slope must be evaluated to determine i this is practical, Surcharged Flow Altiough sewers are notusually designed to flow in a pressurized condition, seme- ‘times storm sewers are designed to surcharge slightly under peak flow. The Maming equation can te used to conduct design analysis of slighty surcharged sewers. The analysis of flow under pressurized conditions in the watei distribution systems and wastewater force mains has been traditionally carried out asing the Fazen-Williams formula, Information on the use of the Hazen-Williams formula cin be found in Chip. C2. Exemple C125. 4 36-in sewer pige (7 = 0.013) # laid on a slope of 2.00 ft per 100) ft. The length of the sewer is 1000 f. Find the level of water is the upstream mashole with respect to the’ sewer invert when 50 cfs is conveyec, Assume that the downstream discharge water level is equal t0 the ciown of the sewer at the dovnstream end. “ Solution. By solvingfor S, the Manning equation undzr full-flowconditions be- cones At Q ~ 50 of; n ~ 0.0:3, and D = 3.0-ft, 8 becomes 0.0056, Note that uader surcharged conditions 5 :n the Manning formula is taken to be the slope of the hydraulic gradent rather than the actual slope of the physical pipe. Therefore for 1000 ft the required grad ent would be 56 ft to convey 50 cfs of Blow. The drop in elevation of the pipe over the 1000 ft at 002 slope is 2.0 ft. Therefcre, the surface of We water at the upstream manhcle would be surcharged 3.6 ft atove the crown of the pipe. + Hycraulic Impact of Bends and Manholes In extensive sewer systems (except in hilly areas), most o! the pipeswill have mild slopes, and the flows will be subcritical. Extra energy losses occur at all transitions ‘where changes occur in size, slope, or direction of the pige and at junctions where several pipes come togetzer. If the transitions aro properly designed to allow for thee energy bsses, the condition of uniform flow may be approsimated in the individual lines and the flows will never be at depths greater than the desths computed on the assumption of uniform flow. However, if the transitions are not properly desigied, then pipes may attimes flow at depths greater thar the comprted dephs and surcharge may ovcur under peak flow conditions. The hydraulic princi- pleginvolved in transitior design are illustrated in Fig. C12.6, which shows a trensi- tion where a pipe flows into a larger pipe laid on.a flatter grade. Owing to the turtulence created by the flow expension, there will be ahead loss /,. To prevent the upstream pipe from Jowing at a depth greater than its norma depth a), theWASTEWATER AND STORMWATER ?IPING SYS"EMS c.641 FIGURE C126 Flow profile x junction, relative elevationsof the bottom of the pipes must be such thet the energy gradient of the cownstream pipe is lower than that of the upstream pipe by the amount of ‘hy This requires that the invert of the downstream be placed below that of the upstreait pipe by.the amount “hp. The relationship betwees the-various vertical dimensions is given by - hip = Hy- Eth (C12.7) ‘The oquation aicumes thatthe loss is soncentratod at the center of thetransition. As actvally constructed, the transition will take place within a manhok, and the channel section within the manhole is made to provide for a gradual transition between the two pipes. If the computed invert d-op hyp is negative, it is usually taken as zero and the pipe inverts are placed at the, same clevation. = ‘The energy loss #, is usually smail, Dut it can assume fairly high values when high veocities ateinvolved. Data on the magnitudes of A, arescarce, but such data as aié sVailable irdicate that A, can be represented asa fradion ofthe change in velocity heads in secordance with the equati s i= a(S ®) (C12.8) where h, is the energy loss in feet (mete's) caused by the transition, K isa constant characterizing the degree of the hydraulic disturtance of the transition, V is the velocity in feet per second, and g is 32.2 ft/s per soond, or it/s? (9.81 n/s*). Values of K for smooth transitions are 2s low as 0.10 for increasing velocity transitions and 0.20 for decreasing velccity transitions. Increased transition losses ‘occur when a sewer line changes direction and zt junctions where one or more branch sewers joit a main sewer. Reliabe information on thetransition head losses in such cases is almost entirely lacking. The hydraulic desigr of junctions may be considered as the design of two or more transition, one for each path o flow. The exit sewer is common to all paths, and its inver: must be placed at the lowest computed clevation. Because of the lack of infornation on the transition losses, allowarces are ustally made in accordance with tke judgmert of the designer. An arbitray procedure which is commonly adopted i to allow about twice as much loss along flow parhs in junctions es the allowances for simple transitions involving the same velocities.cosa MIPING SYSTEMS ASPECTS OF NONUNIFORM FLOW As stated previously, routine design procedures for sewers assune uniform-flow conditions. This assumption, alttough not precisely correct, is adequate in the majority of situations. Note that some situations requir: a more tiorough aralysis 0 nonuniform flow cosditions. ‘The significance of come nonusiform flow conditions is illustrated by reference tc Fig. C127. The flow conditions for each of the cases shown are explained briefly below: 1. A channd witha mild slope discharges freely into the atmosphere (Fig. C12.7a). ‘The critical depth will occur atthe outlet. The depth will increase at successive sections upstream until the nornal depth is reached, beyond which the flow will (e) Note: n= normal depth, de= critical depth, d,_» = ectual flow depth, H = erergy head NIGURE CL. Examples of nonuniform flow profile.t WASTEWATER AND STORMWATER MIPING SYSTEMS c.643 be uriform. In many cases the length of the channel will be much les than the distance required to develop normal depth. The significance of this fact is that for a short strech of pipe, its actual capacity w carry flow without surcharge may 3¢ much g-eater than would be calculated by assuming uniform flow. 2. A chinnel changes slope from mild to steep (Fg. C12.7b) The concitions up- stream of the junction will be the same as for case 1. The depth dowrstream of the junction will decrease and approach normal depth. 3. A channel changes slope from a mil slope to snother mild slope (both slopes steeper than the uniform-flow slope) (Fig. C12.7c). There will be a gradual decrease in the depth from section to section. 4. A channel changes slope from steop to mild (Fig. C12,7d). In this case, the change from the upstream supercritical flow to the downst:eam subcritical flow will tke place suddenly in a “hydraulic jump.” The position of the ump may be either upstream or downstream ofthe junction, dependng upon tie relative values of the various parameters which control the flow pattern, © 5. A steep slope discharges into the atmsphere (Fg, C12.1e),In this cas the flow wil te at the normal supercritical flow, provided thatthe epstream contro has permitted normal depth to be developed. aa . ‘The-foregoing discussion on nonuniform flow is presentec to show the reader the importance of understanding these principles For detailed preseatation of nonuniform flow, the reader is referred to textbooss on the subject? DESIGN ASPECTS OF SEWERS ‘Sewer designs are’ jgovemed by minimun design sandards it many areas. In the United States, ‘various city, county, or stele standards may appiy. Various countries also hav: national standards. Many of these standaris repeat sinilar design criteria. Commenly Used Design Criteria. © = =" ‘Commonly used design criteria include :he following: & ‘+ Minimum size for gravity sewers: 8-in (200-mm) tiameter. . * Minium depth: Sutfcient t0 receive sewage from basements and prevent freezing. * Slope: As recuired to produce velocities no less than 2.0ft/s (0.6 m’s) when flowing full at n = 0.013. Uniform slope between manholes. ‘Sewers with velovities greater than 15 fi/s (45 m/) should be properly protected + Alignrient: Sewers 24 in (600 mm) or ‘ess shouldbe straighi between manholes. + Changes in pipe size: When smaller sewer joins 2 larger pipe, the invert of the larger sewer should be lowered to maintain the energy gradient. This criterion is approximately met by matching tho 0.8 deptr point of both sewers, o: by matchng crowns in smaller pipes. + Manholes: Manholes should be placed at the enc of each line; at all changes in srade, size, or alignment; at all inters-ctions of iewers; and at intervals of notcae HIPING SYSTEMS nore than approximately 400 ft (120 m) in smaller sewers and @0 ft (180 m) in arger sewers, * Relation to vater mains: Sewers shall be atleast 10ft horizontally from water rains, and where they cross, an 18-in (4:0-mm) minimum sepzration shall be maintained. + Design pressure: Sewers convey vastewater under atmospheric pressure. In some instances, vhere the wastewate- must be pumped, tie pipe must be designed for maximum expected internal pressure. For additional infornation refer to Chap. Cl. ‘+ Design temperature: Wastewater and stormwater are generally at ambient temper- atures. The temperatire of the stormwater may vary significantl; depending on ‘he season, such as vinter and summer; however, no special consideration is varranted for the desgn of underground sewer pipes. DESIGN EXAMPLES TN EAMES Design examples are presented for ‘a sanitary sewer‘ system and a storm sewer system in the following section. These examples are adzpted from Ref. & Design Example of Sanitary Sewer Exumple C126. Design a sanitary sewer system for the residential district shown. in Fig. C128. The district is two-tuirds developed; therefore the rrobable future Population density can be estimated without making a detailed population stady. FIGURE C128 Typical map for the design of sanitary sewers. (Alapted from Ref. 8.)WASTEWATER AND STO&MWATER PLING SYSTEMS C645 It is esimated. that the future average saturation populaticn density will be 65 persons per acre. The maximum hour -ate of flow of sewage is estimated at 250 cd. The maximum rate of groundwater infiltration to the sewers, to be provided for, is 2000 gpd per acre of the service area. ‘The minimum size of sewer is to be8 in. The minimum velocity of fow in the sewer when full is to be 2.0 ft/s. The caoacity of the sewers vill be determined by using Manning’s equation with a recommended n value of 0013. ‘Since the homes in this area have basements, the minimum depth below the street sirface to the top of the sewers will be 7.0 fi, (In areas where basements are not normally constructed, the depth of cover to the top of the sewer may be as little as3.0 f.) Solution 1. Draw a line to represent the proposed sewer in each street or alley to be served. Near the line indicate by an arrow the dirzction in which the sewage is to flow. Except in spacial cases, the sewer should slope with thesurface of the street. Itis usvally more economical to plan the system sc that the sewage fromany street Il flow to the pont of disposal by the nost direct (and consequently th: shortest) route. ‘The lines representing the system will often resemble a tree and its branches. In general, the laterals connect with the submains, and these, in tum, connect with the man or trunk sewer, which leads to the point of discharze. 2.-Locate the manholes, giving each an identifcation nunber. 3. Sketch the limits of the service areas for each lateral, unless a sirgle leteral will be :equired toaccommodate an ares larger than can be served by the minimum size of sewer with the minimum slope, "n which czse a further subdivision may be required. Whero the strocts aro laid out,the limits nay be assumed as being midway between them. If he street layout is not shown on the plan, then the limits of the differeat, service: ereas cannot, be deternined as dlosely and the, topography may serve ai a guide., 7 4, Measure the acreage of-several service areas. At this point, the cesign may be represented as shown in the plan view in Fig. 0129. < 5. Prepare a ubulation, such as that shown in Table C12.5, with columes.for the different stepsin the computation aad a line fer each section of sewer.between manholes. ‘This tebulation is a concise time-saving method and shows both the data ard the resuits in orderly sequence for subsequent use. Usecoluran 1 for numbering the linesof the table, for readyreference. Determine. : by inspection the manhole that is farthest from the point of discharge, aint enter its denification nimber in the first line of column 2, and the number corresponding to the manhole next on the line downstream toward the trurk sewer in column 3. Eater the name o’ the street or alley incolumn 4,the length between manholes in column 5, and the area in acres to be served by the sewer ata point jus: upstream up the lower manhole in column 6. ‘On she next lino enter the corresponding data for the nextstretch of zewver, and in colunn 7 enter'the sum of the areas listed in column 6. “he area ir column 7 is the basis for computing the required capecity of the sewer. Enter the data for each section of sewer in the above manner, folowing the line to the point of discharge, including the trunk or main sewer. Enter in column 6 the rate of flow in the sewer which is equal to the maximum per carita rate of sewage flow multiplied by the assumed future density multiplied by the area showa in column 7. Enter in column 9 the rate of allowance for groundwater infiltration, which is equal to the rate 2er acre to be provided for, mutiplied by ‘he area in column 7.cos PING SYSTEMS FIGURE C125 Map showing minholé, Cotumn 10coritains the sims ‘of ih figurés in coluiié 3 and 9, in zillion gallons perday, In coumn 11 ths rate is converted to cubic feet per secoia3, which is the more convenient way of expressing the capacity of sewers, since most diagrims andtables indeate the capacity of arculer pipes in eubicfeet per second (1.0 mgd = 154 cfs). Column 12 contains the requirec sewer sizes; column 13, the slope, columin 14, thevelocity when the sever is full; and column 15, the capacity. Colunn 16 contains the.:levations of the stree! surface atthe manhole corresponding to the identification nunber in column 2. Columns 17 ard 18 contain the invert elevatiors of the upper and lower ends, respectively; of each reach of sewer. «+ In selecting the sewersizes and sopes, the designer makes use 0° Profiles, such a8 the one shown in Fig C1210--This allows the designer to sele:t a minizum sewer size and slope that will carry the computed flow and that 1lso will meet mitimum-depth criteria, Design of Sterm Sewer System Example C127. Design a storm sewer system for the area shown 1n Fig. C1211. ‘The location cf the proposed main storm sewer that is to receive the stormwater fron the distret is showa on the map, and the invert eevation is known at the post where the proposed branch storm sewer is to be connected and for which provision has teen made in the design of the main storm sewer. The required lovest élexation of the invert of the branch storm sewer is therefore known a: the proposed point of discherge into the main storm sewer.“kop apd nd 6 sue et 1B = BOO x SE = ye OL ge ax SE pt SHIM = HOP x 6 | agaseaeas Z i cnaeworsegcaa ae 4 Be 8 suey BoM) sanag Kieyues w 10} suoneindiuoy §ZL9 FTAVL 647cee PIPING SYSTEMS 22 Center Forest Ave, to MH Ashmount St MH Aspen St ‘MH Cedar St Me MH Acorn Si, P= =n eh 75 jn. 180] A: © T0000 7 5000" =90) Disance.t FIGURE C12. Tyzical profile for sanitary sewer design exampk. (Adapted from Ref. &) 4 careful stady of local conditions, including the presont and probable futsre development o; the district, indicates that 70 percent of tke surfacesiin the distiict are expected to be impervious. The inlet time has been astumed to b: 20 min. The stom sewers ae to-be designed for the 5-year level of protection o' the area. For this design example, the rate of rainfall is taken frem the asstmed curve of intensity of pricipitation represented by the formula i = 20.4/¢9, in which 1 is rainfall intensity in inches per hour ard ris rainfall durationin minutes This formula Teprisents the average rate of reinfal. for a duration of t mia which may be expected to be equaled or exceeded on the average once in a S-yearporicd. Te rainfall end runcff curves are shown cn Fig. C1212. : Note that the intensity of precipitation curve in Fig. C12.12 is site-specific ior the geographicel area under consideration. For different areas different curves may exist, and they can te obiained nornally from state agercies. Note also that the intersity of presipitation ia the example is for the 5-year xecurtentiaterval storm. For ihe design of systems ‘or greater or lesser levels of prctection, diferent curres would be used. The runoff curves ate provided for various coefficients of runoff C. The curves alot out the Ci terms in the rational mettod equation Q = CiA. Units for runoff are in inches of runoff per hour. . To illustrate the use of the chart, assume that a particular area as a time of‘VASTEWATER AND STORMWATER PIPING SYSTEMS cea FIGURE C1241). Nap for storm sewer desigr exanipl : +. ‘concertration; of 60 mift and that the area‘is 70 percent isipervious. The runoff would be read drectly from the Ci curve for:7 percent inpervioustess at 1.02 in/h of-runoff: Multiplying this-number by thé crainage ama-in acres yields the runoff flow in cubic feet per minute (1 cfs is almost exactly equivalen:-to 1 acre- in/h). Note that 2°70. pereent impervious. area yields a coefficient of ianoff C of ‘about 0.60 for a 6)-min, duration storm The weighted-average coefficieat of runofi Cof an area composed of 70 percent impervious area at a rmoff coefficent of 0.80 and 30 percent pervious area with a ruroff coefficent of 0.15is 0.605. Runoff from .60-acre area with a composite C of 0.05 for a stcrm with ai intensity of precipita: tion of 1.68 in/h vould be 61 cfs. This result can ilso be obiaindd by reading 1.02 in/h of the 70 percent impervious curve on Fig. C12.12 for a 60-min duration and multiplying by the area of 60 acres, to yield 61 ef. While it was recognized that storm sewers cesigned o1 this basis might be ‘overtazed on the average of once in abcut 5 years,it was not considered reasonable to provide for sterms of greater intensity, because of the greater cost. During the earlier years of tte life of storm sewers, they will be able tocarry the rinoff from higher rates of rainfall than they will te able to carry later, because the assumed coefficents of runoff are based upon future (with larger fraction of impervious areas) rather thanpresent conditions, which may include a larger fraction of pervious areas. A progrosve iterease in impervious surface and ass2ciated ruzoit will be caused by the gradual substitution of :oofs and aved areas for presently unim- proved areas, Figure C12.11 shows the drainage area. Street elevations axe shown inFig. C12.8. ‘The limits of this area are influenced not only by the surface contours Dut also by(9 fou wo1f paidepy) “ssousnowsdas yo s20:82p mousen 20) sonino Aysusnu won aag sd 322 =, vonaycaad jo sven] aw anow repay i nyu cron yepie c.650WASTEWATER AND STCRMWATERPIPING SYSTEMS, C651 the service areas of existing storm sewers. In a district where the surface slopes are ‘moderate and generally uniform, contour maps may not be required. Instead, surface clevations may be adequate if they are obtained for street intersectiors, for high and lov points, and at locations of change of surlace slope. ‘The storm sewers are to be designed, in general, with the crown ata depth of at least 5 ft (1.5 m) below the surface of tho stroct. Tho minimum size of sewer is to be 1? in (300 mm). The assumed minimum average velocity is 3.0 ft/s when fiow is at ful depth. ‘The capacities of the sewers are to te determined by using a value of n= 0.013. ‘Velocities for flow at design: condition: may be tigher for dorm sewers than for sanitary sewers because the design storn flow is many times greater than the peak sewage low in saritary sewers, Because of high velocities, iti importantto provide for addtional head to compensaté for losses, such as those due to bends, manholes, transitions, ‘and velocity changes. Solstion. . 1, Draw a'lineto represeat the storn sewer in zach streetor alley tobé served. Place an arrow near each sewer, to shov'the diredion:of floy. The sewers should, in general, slope with the sticet surface. will usually prove tc be more esonomical, however, to lay oat the system so that the water will reach the main storm sewer by the mést direc: route. wes - ; 2, Locate the manholes tentatively, giving to exch an ideatification number. In this example,-a manhole is.to be placed at each tend or angle, at all junctions of storm sewers, at all points of change in size or spe, and at intermediate points where the distance exceeds 400 ft (120 m).on 12-.t024-in (300- to 600-mu) sections. Where a:good velocity would:be.maittairied dung practically’all conditions of flow, and the sewer is large enough for workers towaik withcut stooping, intervals between manholes up to 600 ft (180 m) may be wed. Sufficent manhcles should be buil: to allow access fot inspection and. cleaning: Later, when the profiles are drawn and the firal slopes are fixed, it may be desirable to change the locations for some manholes so that the sewers would,be at the most advantageous depth, Particularly where the slope of the street surface was not substantiall7 uniform. ‘Other considerations, such as obstacles underground, may require the installation of additional manholes, dué to change in alignméni or special forms of ccnstruction involved in;junctions or connections with other sewers. a 3. Sketch the limits of the drainage areas tributary at each manhole. The assumed characteristics of future development and the ‘opography will determine the proper limits. 4, Measuce each individual area by slanimeter or other methods thet will give equally satisfactory results. Now use of computer programs nakes it mich easier to obtain very accurate results, : 5. Prepare a tabulation to record the data and seps in the omputaticns of each section of sewer kctween manholes. ‘The computations for a selected line of this section are shown in Table Computations are carried out as follows: Columns 1, 2, 3, 4 from the layout map of the system in Fig. C12.11 Column 7'is the travel time to the upsream mamiole considered. For Ine 1,20.0 nin is the ialet time for overland flow to reach the farthest inlet. For lin: 2 it is 200 min plus the travel time (1.7 min) in the sewer from poiats 1 to 2. The travel time in section (column 8) is calculated by dividing column 4 (length’ by colums 13 (velocity) and then dividing‘ur = groc0 xy S»SSo] puag pus aseanur prog Lypojas Jo} moll pIMOM JoMSs YO UES yp 7] uumnjen ur sain 2p 44 p ucanjoo wy asowy Busta £9 pa ur a10u1 Suicinyw Aa pourERIO ae OT MUIIOD ur SaINdA “EL GuNIOo ur sandy oun Ka BE A Aq 348 g wanyO> ut sanBly eLSBL CELL OST OBE ‘SORE OLESI S90 SHE ovest zoom ceo Cute Tost OLD6I GEO ee 09061 ELTol sco gee. eU16I Wist FTO Tae LEGE STO OTD OTK crest oes cso Sue ‘9000 8 OFT “1000 99 oer plod 99 Ove Fran 09 riz Sio00 09 GVO am00 Fs cre sooo ps e3L. 100 BF £79 zoo zr OW so Puwarena Br ose ass UE Sa -*BsaiwD ST cor W849 SE ose = asuaIwD FL gIe zm 95, se ISB er $6 “ys ZL oy ise TE oe "Sam OF ow se oor Es LEESL LED eee 91000 ZF ODF o@ se zor ass 6 Lepst Isher yO eee ad Bw C6 09 aga31m 8 Isvol SeSGI ¥O TZ of ear Te ce re asa oser Sesct SLO vLKe a on e% FOI cl SE (TS PIMPY 9 O996t SOueT SHO Ute ve 6m ox ry Sl Sa SOLE SOUST OSD OME an ve 69 tom WAN O6cst osest oso Se wom. ee cy ye Sor € 66 OSes 060 HR st oe Ow et et me NZ @o aw wo ew a OD ow” oO w& © @ pro pus yea "won ae uM same some wonsol OL, snot sed, “ona see spe mer Sem “yu —__ aun aig eles soddn, a0 "WNT HOUT arin er. me ——— aun para ssroneg wu0ig 40} suonerediwe) 9°ZL9 TGV C552WASTEWATER AND STORMWATER PIPING SYSTEMS cess the result by 60 to convert the time to miauites. Coltmn 9 (runoff rate) is read from Fig. C1212, using the 70 percent impervious curve for the tims listed in column 7. Column 10 (required capacity in cubic feet per second) is caleulated by multiplying column 9 by columa 6. Columns 11 and 12 involve designing the pipe segment. A particular size and slope are determined from the Manning equation Slope are determined by the constraints of (1) maintaining adequate cover of 5.0 ft over the crown of the pipe at upstream and downstream manholes, (2) maintaining « minimum full-low capacity of 3.0 f/s and (3) avoiding as much as possible excessive excavation depths. Column 13 (velocity) is determired from the Manning equation and from Fig. C125 for deterinining velocitics in partially flled sewers. Columa 14 is the full-flow capacity o: the pipe as determiaed from the size and slope. The capacity should equal or exceed the required capacity listed in column 10. Column'1S (suriace’ elevation) is taken from st'eet maps, survey information, or topographical maps in preliminary cesign.” Column ‘16 (fll) is calcilated by ‘multiply'ng column 12 (slope) by column (length). Column 17 (invert at upstream end) presents the invert at the upstrean manhole. At the farthest manhole the required depth of cover of 5.0 ft normally governs as a starting point. Note that in line 1 the invert of 199.90 plus the pipe size of 1-3) ft (15-in pipe with allowance for pipe thickness) gives an elevation of the top of the pit ft below the surface. = Cotunn 18 (invert clevation at dowastream erd) is calcilated as colin 17 (upstrean invert) minus column 16 (fall). The lower or downstream elevation should be checked to enstre that 2 minimum of 5.0 ft of zover is mintained. Note that when pipe sizes change, the crowns of the seweis art generallymatched. “herefore columa (7 on line 2is 050 ft lower thaa coluina 8 for line l, This is eecounted for by the incvease in pipe size from 15:0 21 in. ach ateral is ten designed in similar wey. I aecessaryte fst design ot the submainis subsequently modified so as toserve the literals properly. It is possible in some cates to’omit some inanboles on latéial sidrin sewers, Usiag the inletsubstruc- tures at which junctions and changes in size, direction, or slope may be made. In relatively flat areas,:the design it influenced by the: ‘otal availeble (clevaticn difference) between the outle: of the aystem and tke upstrean area. In such cases, the designer neéds to be cognizant of the overall limtation when starting the analysis. In sucn areas, larger pipe sizes at flatter slopes vill be required. Finaly, the outet conditions at the downstresm end of the system maj be variable, especially if the storm sewer discharges to a stream o1 river. In such cases, the designer must ceterminé an outlet design condiion (level of the rive) to form ~ the basit for the hydraulic computations The outlét conditioa, which may be the ‘5-year recurrent staze level in the river, is determinec in conjundion with policymak- ‘ers. The outlet coniition selected will, especially in flat areas, have a large impact ‘on the design and the cost of the projec. SEWER PIPE Availabe Pipe Meterials The macerials of vhich street sewer pizes are myst commenly constructed are vitrified clay pipe, concrete, plastic, and iuctle irox pipe. Suggested specifications and other pertinen: information about sswer pipes are preserted in Table C12.7. 201.20. This is 5.0.”doo0e yIof 20} 40119 HUIS ‘unsay uonenye UE UOHTEsIUS! 305 6960 IS ‘Bunsen se 30} 726 HLSW ftwo u €2t ‘oat ‘sus ‘sc pe=y ‘Fupsor ap 305 9269 HLISY wo Sunsa1 aonenn2 adyd Ieandie jruozxoW 20) 11ST « “yor 01 da somes: vwohids 7 119q"spurg Buy -jeos feusows 2180 WISV ‘sa3p08 Jogqns ‘ef sete a we ePID 29508 soqqna ‘uteh solids » nea ert0 WIS soype8 Jaqqn “ict cox ML ovez xPE £050 WASV suzap monde saxowro-va3i0haast aL ‘t-ar ’ one eo ras a ott 9 Wis hd aaowos-pamiofuay, Jamas Cyceid 10} pu BONEAATE 30} 96 SY YEO} sor dn taMas— oHde FHA EMO WISV AL ST TONEY ‘uysan 3 30} 7260 SV ‘uy § 20) ndop porns, PA OPH 9D wip WISY oid spew ops) woneny> jenwems fuyroouiue adie pow wo ‘gayeed sana af oRyds-pur-qog seo WUISY west 019 WIsv sao wisy’ oy ay (0009 LIS 2a up poatunn ‘Sueuo ‘suoneouods inde suonesyioad wind) yore = awoneeyads ‘One paling 3 eon epiouue yeplourioy —- jousteal ad pad wor swulor pue 2uid)q 10) suorwoyrods payseBng pu voneuoyul [eID £ZLOAIEVL 54 csSie saqgns ‘Suey Sod oqueyosur . s08Kes 9 Ueq TT'TCY INV 0Hids 9199 Libd WISY a3 ‘URC NASY om 60827 Samos Ames 30g 199s Ayea8 304 sonas Anan poe ‘did atg 07 dn so “ayBooats somes pow. a pmrOpUY cess,cose >IPING SYSTEMS Vitrified Clay Pipe. Vitsified cay pipe is manvfactwed in stardard and =xtra- sirength clasifications. It is widely used, especially in smaller sizes because of its resistance te corrosive wastewaters. It is generally available in sizes from 4 in to 36 in (100 mm to 900 nm), in laying lengths from 1 to 10 ft. Larger-diameter (48 in) pipes may be available by special order. ConeretePipe. Concrete pipe is available in either plain or rénforced chssif cations in various strergth categories. Concrete pipe & widely ured especidlly in larger sizes. JInreinforced coneret: pipe is available in sizes from 4 in to 36 ia (100 mm to 900 mm), and reinforced :oncrete pipe is avaihble in sizes from 12in to 144 in (300 mm to 360mm) in five strength classiicaticns. For details of clasifica- tion refer to ASTM Standard C76, Joints are normally made with rubber giskets in grooves formed in tke tongue. Other jointing systens are also available. Advantages of conacte pipe are its wide range o! sizes, layng lengths, and stengths. A disadvantage of concrate pipe for sewers is that itis subject to corrosion under acidicconditions. If flow velocities are insufficient to prevert the deposition of organic sclids, septic conditions may result. Hydrogen sulfide gas produc-d by araerobie decompositicn of organic matter becomes oxidized to sroduce sulfuric acid, which damages the pipe. This concition can usually be prevented by designing the sewers so that self-cleansing wlociiies will occur rrost of the time. Protective linings including coal tar, coal-tar epoxy. vinyl, and epcxy mortar: can be wed to prevent corrosion where wastewater is expected to be highly acidic or where deposi- tion of solids is anticipated. Additional information «an be oblained fron the American Concrete Pipe Association. Plastic Pipe. Plastic pipe used in sewago systems inclades PVC (polyvinyl chlo- ride), ABS (scrylonitrile-butadiene-styrene), and PE (polyethylenz). All offer ad- vantages of corrosion resistance, low friction characterists, and light weight. Plastic ives are generally less rigid: They require proper bedding and lateral cugport. Palyethyleneand ABS pipes are abo available in a variety of standard and preprie- tary product, The reacer is refewred to manufacturer: for more informaticn on these pipes. Available size ranges from 4 in to 15 in (100 mm to 275 mm). Ductile Iron Pipe, Dustile iron ripe (DIP) is employed in sewerage primarily for fo:ce mains and for piping in and around buildings. It is generaly not used for gravity sewer applications. APPURTENANCES AND SPECIAL STRUCTURES Exential to all sewerage systems are the appurtenant structures such as secvice comnections, manholes, unction clambers, stormwater inlets. and diversion cam- bers. The design of such structures is not covered in detail in this chapter, but typical desigrs for the most commonly used appurtenarces are presented briefly. Building Sorvioo Connections Figure C12.13 shows typical details of service connections to a sanitary sewer laid ina relatively shallow trench; Fig C1214 shows a typical connertion to a teep sewer. Note that the connection siown in Fig. C12.13 makes use of either a wye branch or a tre branch in the main sewer line.WASTEWATER AND STORMWATER PIPING SYSTEMS cos pours ne 40" en . 2 ' TE we branch For fuwre connection plug this point 1A) Pian ‘Mii sewer Juncton Chambers and Manholes. Figure C12.15 shows a typical design for a. junction chamber.and:mankole for relatively small sewers, For manholes cn large Severs, a spedal undergiound struc- ture will ordinarly be required, and the entrance to it will be provided for by « manhcle located at-one side. Such chembers anc manholes are required at every sewer junction ard at every point where the sewe: changes in size, slope, direction, orelevation. Itis general practice to install sewers i straight lines betwee manholes, except that for the larger sizes (36 in (900 mm) end’above] which may be laid or carves Manhole: are usually installed at the tipper end of every lateral sewer anc in straght-line sewers so that the spacing ‘will not exceed about 400 ft (120 m) te 600 ft ‘180 m). Figure C12.16 shows typical detai of a “drop manhole’ at a point where a sewer takes an abrupt drop ir grade. Stormwater Inlets Stormvater inlets which carry stormwater from the streets to the storm sewers are located upstream of the crosswalks at street intersections and at low 2oints. The designs vary considerably, and most cites heve adopted their own standard design, details. There arethree general types o' inlets: (1) curb inlets, which have a vertical opening in the cirb; (2) gutter inlets, in which horizontal opening ir the gutter is covered by a eist-iron grating; and (3) combination inlets which combine both the above features. Many types and sizzs of standard castings are available for the constriction of irlets,ose, PIPING SYSTEMS déertiture EPS ‘oanecion pi ttn pel Section AcA i Section C-C FIGURE C12: Junction chamber and manhole for small sewersWASTEWATER AND STORMWATER PIPING SYSTEMS C659 co. Section Aca FIGURE C1216 Diop manhole, > Sewage Pumping - In many sewer system layouts, it is nezessary to provide for pumping at one or more points. The required pumping capscity will vary from a few gallons per minute for stations serving only a few laterals to many faillions of gallods’ por day for stationsserving large districts. The smaller stations are frequenly built uncerground, either 2 built-in-flace or as complete factory-assenbled units, Pumping is usually done wth nonclog centrifugal pumps, although pretimatic ejectors aré :ometimes: used fo: the smallr installations. For detailed disxussion of the design of sewage pumping. stations, see. the ASCE-WPCF manual." = Structural Requirements © > ns : Sewers must be insialled so as to be able to withstand the loads imposed upon them by the weight of the carth and any superimposed leads. The supporting strengtk of a buried pipe depends upon the instalation conditions as vell as the structural properties of the pipe itself. Sewer pipes are clasified as rigid pipes, which cannot deform materially withou: cracking. For rigid pipes in treaches, the load cau be represented by the equation W =CwB* (C12.9) where W = load, lb/ft (kg/m) of length 5 v = weight of soil, Ib/ft? (kg/m? : 3 = width of trench at the top cf pipe, ft (n) C= dimensionless coefficient wiose value depends upon type cf soil and ratio cf depth of cover to tiench widtt Table C12.8 gives values for C, and Table C12.9 gives values of w to be used in the equation, If a pipe is placed on undisturbed ground and covered wih a fill, the load can be estimated from W =CwD? (C12.10)HPING SYSTEMS TABLECI2.8 Values of C for Use in Formula W = CwB? Ratio of depth to tench Sand and Saturated“ Damp ‘Saturated vwieth damp topmil___tepsoil_——_clay clay 05 046 036 ‘oar oar 10 08s 086 088 0.90 15 118 11 124 128 20 146, 130 156 ve 25 +70 176 184192 30 190 98 208 2.20 35 2.08 a7 230 2a 40 22 233 249 2.66 45 234 aa 265 287 50 245 259 2.80 3.03 55 254 269 293 319 60 net 278 308 333, 65 2.68 2.86 34 3.46, 0 23 233 322 357 75 278 298 330 367 80, 281 303 337, 316 85 2a5 307 342 3.85 90 288 ait 343 3.92 95 290 haa 352 3.98 109 22 317 356 oe nD 298 Sa. 3.6 414 129 297 324 368 422, cE) 299 an an 425° > 3.00 328 375 434 15) 301 330 377 438 Very sreat 308 333 3.85 455 Sourc: Jowa State Univ. Eng. 2xpt. Sta, Bul. 47 TABLE C12.9 Weights of Ditch- Filling Materials Materia, torte Dry sand 100 Orcinary (damp) sand ns Wer sand 120 Danp clay x20 Saturated clay 130 Saturated topscil us Sani and damg topéci Wore: Toit? « 1602WASTEWATER AND STORMWATER PIPING SYSTEMS c.68t TABLEC12.10 Values of C for Projection Condition: — eee ee Ee Ratio of depth of cover to pipe diameter 05 10 15 20 25. 30 35 40 c 06 12-20-3042 St where D is the pipe diameter in feet. Table C12.10 gives values of C fer the latter condition, which is known as the projection conaition. The load on a pipe placec in a trench will :ncrease with the treach width until it equals the load for the projection conditon. If there is a doubt-as to whether the ditch condition or the projection condition controls, the load should be calculated by both formulas‘and the maximum yale used. TABLE C12.11 Percentage cf Wheel Leads Tranismitied to Underground Pipes for Unpaved Roadway or Berm Areas (Tabulated iigures show percentage of wheel load applied to 1 lin. tot pipe) . com Depth of Trench width at op of pipe, it acti, over tp of : : aos 4 5 6 7 170 26.0 28.6 -°29,7 29.91 N2:.303 - B38 162 ABS 27 MB VDI BO ate WB ABS 6 14.8: 71588 16.7 oe ey 72-90. 103 ‘15-123 : 33 5.06..635 5 -73-; .335> 9.0 4 Give MW 2337 GMT SBn 52 TO a I St 70 Ti IGads ranted ae pracaly neglighle below 6 “ sno DoH pecenaper ined beth ive lod nd impact tamed to Pipe In aidition to :he load of the baekfil, some alowance should be mide for the superimposed loads caused by vehicles. [tis usually safe to asiume that H-20 wheel loads will be the greatest live loads to be suppocted. H-20 loads refer to trucks having a gross weight of 20 tons, 80 percent of which is on the rear axle, each rear wheel carrying 8 tons. Table C12.L1 gives the pewentage ‘of wheel loads that can be assumed to be transmitted to buried pipe? Pipe Bedding Conditions ‘The supporting strength of a rigid pipe depends uoon the type of bedding used in the installation of the pipe. Four genezal types 0: bedding conditions have been defined for ditch conduit Type 1: Impernissible bedding. Littleor no careis taken toshape the foundation to fil the lower part of the pipe or to fill and temp around the pipe.c02 PIPING SYSTEMS Type 2: Crdinary belding. The soil at the bottom of the trench is shaped to fit the lower part of the pipe witt reasonabie closeness for a width of at least 50 percent of the pipe diameter; and the remainder cf the pipe is coverec to a height of at least 6 in (15 em) above its top by granular materiel which is 1and- placed and -tamped, Type 3: First-class bedding. The pipe is carefully bedded on fine granular meterial in, an earth foundation carefull) shaped to fit the bottom part of the pipe for a width at least 60 peicent of the diameter; the remaiader of the pipe is entirely surrounded to a height at cist 1.0 ft (30 om) above the wp by granular materials placed by hand in layers not exceeding 6 1n (15 em) and thoroughly tamped, Type 4: Conerete cradle bedding. The lower part of the pipe is embedded ir con- crete. ‘The load factors, or the ratios of the supporting strength to th: crushing load, as determined by the three-edge bearing method (ASTM Methods C497) for the various types of bedding are genevally taken as follows Inpermissible bedding 44 Ordinary bedding 21S . First-class bedding 19 ‘Concrete cradle bedding 2.2-3.4 The factors for the conerete cradle bedding depend upon the amcunt and quality ofthe concrete used. The value of 22 will generally apply wher the concrete exends from about one-quarter of the pipe diameter [with mininum of 6 ir (15 em)] telow the pipe to the height where the lower 120° sector radii ntersect the outside of the pipe. If the concrete is carried up to cover the entire bottom half of the pipe, the loid factor may be as high as 3.4. If the entire pipe is encased in concrete with a minimum of ).25D (4 ir, or 10 cm) both above and below, the load factor may be as high as 4.5. Sefety Factor ‘The specified minimum strength by the three-edge bearing method for a rigid pipe should be divided hy an appropricte safety factor in orter to obtain the working strength. Sone engineers use safety factors as low as 1.0 to 1.2 for reinforced- concrete pipe culverts. For street sewers a safety factor of 1.5 is recommended by the ASCE-WPCF Manual. OPERATION AND MAINTENANCE Protecting the investment in sewer systems requires attention to installation, testing, opzrations, maintenance, and cleaning.WASTEWATER AND STORMWATER *IPING SYS"EMS c.663 Start-up; Operation, and Maintenance Senitary arid storm sewer systems should be cleantd and tested prior tc being put into service. For bath types of sewer systems, tests should be carried outto confirm the follswing iter: * Alignment * Elevations and pipe grades + Watertightness of joints * Location and acessibility of manholes Relevant data should be used to:prepare a sewer tlas which will includ: locations and elevations of all sewers. This atlas is generally crucial to good operation and mainteaance of the sewer system, ‘After the sewer system.is put into use, adequate care and maintenance must be provided to ensure continued good operation: Preventive maintenance vill include * Spot inspection: of sewers for damage at manholes * Identification ard correction of system misuse sich as illegal connections + Identification of missing or damaged manbole covers. * Inspections to icentify accumulations of grit and grease andthe intrusion of roots into the system Pipe fe joints ahdild be’ teeta for ‘wistertightieceusirg an'air test or a water tel. Air tests should be’cartied out in accordance with AS7M C924. Water testsshould be infiltration tests where the deptir of groundwater is'sufficient or exftration tests where groundwater is insufficient to'submerge the section of tie sewer tc be tested. No standard testng’methods are availabe forinfiluation/exflration tests; but local ‘gencicsgenorlly have maximum allovable rates Inspécions Periodically, the insides of sewers should also be inspected. For sewers too small to ente;, inspections are done by using TV cameris which are pulled through the sewers between manholes. Modem televising systems can eater sewer laterals as small as 4 in (100 nm) in diameter and :elevise mains in’ excess of 60 in (1500 mm) diameter. The lateral inspection camer: can gain access throgh a sewer clesnont ‘orcan be inserted irom the main with the use of spezial lateral aunching equipment. For larger sewers, inspections can also be carried out by wading or floating through them ina boat. The frequency of such inspections vill dependon the type of system (sanitary or storm); the pipe material ard joint types; and the local experience with the system. Inall sewer inspections, safety is a mzjor concera, especially with sanitary sewers where there is a dinger of septicity, Prezautions should be taken to avoid exposure to dangerous gases or to the absence of oxygen. Adequate ventilation must be provided, and rules must be established to indicate the approfriate cours: of actionB64 PING SYSTEMS in emergencies. In addition, all Incal safety requirements must be satisfied, and personnel should be given adequete training and equipment. Cleaning Based on results of insvections, sewers and appurtenances such 2s manholes and czich basins must be deaned. ‘Sewers can be cleaned by pulling, thrustirg, or dragging some form of ixstrument :hrough them. Clcaniag can also >¢ accomplished by flushing the sewer, using a sudden rush of water dows the sewerat high velocity. Tre most conmon sewer cleaning systems are the hydraulic jet, sewer rodder, and backet machines. Sewer cleaning programs are underteken to provide cleaning of allsewers generally once every 3 tc 5 years. More frequent cleaningmay be required in sewers where low flows or flat slopes do not provide self-cleanng velocities. On storm sewer systems, itis important to clean catch basins which are designed tointercept tettleable solids before they enter the sewers, Catch bisins are cleaned by hand, suction pumps, or grab buckets. DISPOSAL OF WASTEWATER AND STORMWATER Stormwater zan otdinacily be disposed of by discharge into any natural drainage ctannel. Hovever, increased emprasis is being placed on water quality impacts of stormwaters on the receiving streams. Storm sewer sysiems may require inchision ofsediment traps, wet ponds, and cther features which reduce the pollution loadings toreceiving streams. Sanitary sewage and industrial wastewaters containing ebjec- tionable constituents must be disposed of in accordance with the requiremeats of the local health and environmental authorities. The most satisfactory method of dsposal of sanitary sewage is to convey it to an adequcte public sewerage system. In areas which do not nave publi: sewerage systems, individual disposal systems must be provided. These will vary in size from septi: tank systems, for private residences tolarge treatment plants handling the wastewaters from ‘arge institutions and industries. The design of such systems is beyond the scope 0! this hand>ook. See Ref. 1 for further information. REFERENCES 1. Metcalf & Eddy, Ine, Wastewater Engineering: Treatment Disposal and Reuse, 28 ed, MeGraw-Hil, New York, 1991 2. Design and Construction of Sanitary and Storm Sewers, Practice-No. 37” (“WICF Manual of Practice No, 9”). 3. Great Lakes-Upper Masissippi River Board of State Pubic Health and Environnental Managers, Recommended Standards for Wastewater Facilites, 1976 edtion. 4, Frequency Distributions and Hydrovlimatic Characteristics of Heavy Rainstorms in Hinois, Bulletin 70 Mlinois Stare Water Survey, 1989. - ‘SCE Manuals of EngineeringWASTEWATER AND STORMWATER PIPING SYSTEMS css, 5. TR, Canip, Design of Sewers to Facitate Flow, Sewage Ind. Wastes, vol 18, 9.1. 53 January 1946 6. Ven Te Chow, Open-Channet Hydratlics, McGraw-Hill, New York, 1959 7, E.W. King and E.E. Brater, Handbookof Hydraulics, 5th ed, McGraw Hill, New York, 13, & Metcalf and Eddy, Inc, Wastewater Englaeering: Collection Treatment and Disporal, NcGraw-Hiill New York, 1972. 9. Clay Pipe Engineering Manual, National Cay Pipe Institute, Lake Geneva, WI, 1995.CHAPTER C13 PLUMBING PIPING SYSTEMS Michael Frankel, CIPE Prisident Utility Systens Consulants Somerset, New Jersey ‘This chepter provides the necessary criteria to snable accurate pipe dizing for various plumbirg systems and cost-effective selection ‘0! appropriate plumbing piping materia.:? . Plumbing systems directly affect the health and safety of thepublic, and thusthey are distinguished fom ther piping systetis by the following general requirements! 1. The design, materials, and installation of the systems are directly regulated by 18 Pham oe cian atoning inns ee et cube 2, System design must be ‘approved by an authorized code. sfficial charged with the responsibility of ensuring plumbing code compliance. 3..A permit for irstallation. of the systems:must be, obtained from the authority oo having jurisdiction, e415. 6 Lang 2 4, The systems shall be installed by an individual duly licensed by the authority ” having jurisdiction tor determining the competeace of an individual to obtain a plumoing instalation licese. This mty not be required in some jurisdictions. 5, The installed systems are required to be inspectec and appréved by an suthorized code official charged with the responsibility of code enforcement. ‘The basic plumping systems are 1. Sanitary drainage systems 2. Sanitary vont systems 3. Stormwater dranage systems 4, Potatle water systems 5, Fuel gas systems CODES AND STANDARDS Plumbing codes establish minimum accertable stancards for the design ani installa- tion of tie various plumbing systems and for the components that comprise them. 367c.068 PPING SYSTEMS ‘There are five plumbinz codes thit are generally accepted in lame areas of the country: 2s Code title Sporsors AA) Safety Requirements ‘or Cosecretariats are Plumbing * Plunbing, Heating, Cooling Contractor—National Association + Mechanical Contractors Astociation of America ‘International Association of Plumbing ind Mechancal Officials International Pumbing Coie Intermtional Code Council comprised of the Building Offtials and Code Administrators, International, the International Conference o1 Building Cfficials, and the Southern Building Code Cengress National Standird Plumbing Plomting, Heating, Cooling Contractors—National Associ- Code ation (Om: and TwoFamily International Code Council cemprised of the Building Dwelling Code (soon to 0 Offfeials and Cede Adminisratorr, International, the «galled the Infernational International Conference of Building Officials, and the Residential Code) ‘Southern Building Code Congress Intemational ‘Unform Plumting Code Intermtional Asscciation of Plumbing and Mechanicel Officials " Some states and larg: cities have adapted codes septrate from these builling coces. Because of this nanstandardization, the plumbing code used for each specific project must be obtained from a responsible code official. Many of the tables and figures in this chapter are used only to illustrate and atigment discussions of tre system sizing procedures and design methods and should notbe used for actual design purposes. The information pertaining to those systems is included in the approved plumbing code and should be the primary criterion for accepted methods and sizes for use on any project. There are many nationally recogrized standards that es-ablish dimensions, minu- facturing methods, material composition, tests, and numerous other details specific to individual components of the phimbing system. A. partial list of organizations originating such Standards adapted by various plumbing codes is as follows: American National Standards Irstitute (ANSI) American Society of Mechanica Engineers (ASME) American Society of Sanitary Engineers (ASSE) American Society for Testing ard Materials (ASTM) American Water Works Association (AWWA) American Welding Society (AWS) Cast Iron Soil Pipe Irstitute (CISP1) National Fie Protection Association (NFPA) National Sanitation Foundation International (NSF1) Plumbing nd Drainaze Institute (PDI) Underwriters’ Laboratories (UL)PLUMBING PPING'SYSTEMS c.669 FIXTURE UNITS The fiewre unit (FU) is an arbitrary, comparative value assigned to a specific plumbirg fixture, cevice, or piece of equipment. FU values represent the probable flow thet fixture vill discharge into a érainage system or uie (demand) from a potable water supply system; compared to other fietures. ‘The 1se of fixture units for plumbing systems was expanded by the lste Roy B. Hunter, of the former Bureau of Siandards. In the years since the development of the Hurter method, new fixtures, changes in the patterns of water use, and water conservation woul!’ now result in overszed water and drainage piping systems if ‘the orignal criterie were used. Long-term data andmodern stitistical methods and analyses have resulted in revised figures which we-e used in calculating the tables ‘and charts provided for this revision of the handbook. Since sanitary discharge and water demand FUs are different, the designations DFU for drainage ixture unit and WFU for potable water fixture unit wil be used to differentiate beween them. PLUMBING FIXTURES: : A plumbing fixture is any approved receptacle specifically designed ‘o receive human and other waterborne waste and discharge that, waste’ directly iito the satitary drainage system, usually with the addition of water. Heal fixture materials should be nonabscrbent, nonporous, nonoxidizing, smooth, and easily deaned, Plumbing codes usually mandate’ the number and type of ixtures thet must be provided for specific building use, based on the proposed population. >rovisions for the aandicapped have been made ar integral part of code requirements, man- dating tie nuniber, layout, and barrier-iree access to those fitures. Potadle water discharged from speciic plumbing fixtures may be restricted to a maxinum flow rate mandated by water conservation requirements’ Refer to specific code provi i ‘Table C13. lists average drainage ad vent DFUs, hot and cold water WFUs, gallons-ser-minute (gpm) flow, and braach size information for typical ixtures. EQUIVALENT LENGTH OF PIPING r ‘When one is calculating the pressure los: through z pressurized piping system, one of the fictors to be considered is the eeuivaient length of pipe. This is the actual pipe run plus an additional length, expressed as a aumber of foot of strcight pi that would have the same friction loss as that occurring through various fittings, valves, and so on. Figure C13.1 gives the straight cun of pipe for both water and gas systams equal to various valve types and fittiags for different pipe sizes. An often used and generally conservative method of quickly finding the equivalent run is to add 50 perceat to the actual meastred pipe ran. PIPING MATERIALS: GENERAL When materials and jointing methods ar> chosen fer use in a project, the plumbing code usially specifies pipe materials pemitted to 9e used for the various systemsc.70 FIPING SYSTEMS TABLE C13.1. Typical Plambing Fixure Schedule Drainage Water Sze Stee — rap vent Sie Size Flow, ture ype NeT NPT edd hot ‘Automatic clothes waster 2 a o% Babroom group(WC, LAV, ‘SHIBT) FV Bathroom group (WC, LAV, 6 SH/BT) tank Babtub (GT), wih or without SH. 2 Me 2 & wm os Bicet 1° 4 ™® 1 & 2 Cis sink 6 3 Mm 2 w w 3 Dishwasher, domestic 20% Mh 1 ® w 3 ‘Deatal lavatory, zuspidor, and unit 1% M J & woo Drinking fountain wom mM ew % Floor drain 5 3) Ih Kithen sink andiray, with focd grinder = $2 MHS Kivhen sink and tray, single 14. trap 2 Mm % 2 w w% 3 ‘Kithen sink andtray, muliple 1¥einiraps 3M MDS Lavatory. private 10M mM 1 ® % 2 Lastory (LAV), public 20M mM 2 h %® 2 Laundry tray, on or to compartments 2m Mm 2 wk ws Shewer (SH) perhead ot stall 202 MM 2 w w 3 Sine service (SS) trap xandard 2°30 MH 8m me Sint service, P inp 20 MH Mm 2 wow 4 ‘Sint pot and seulery 2 M2) 2 ww 45 ‘Sint bar mm Mw MH 1 ww 2 ‘Sint using im 63% 5 4 Sink, surgeon's 302 Mm 2 ke Sint wash fountain, per faucet 2m Mm 2 % % Uriel pedestal, Nowout 6 3 m wm a ‘Urinal washout 4 2 MH 5 & ‘Waer closet (WC), prvate Hum 6 3 Mm wo oa valve (FV) ‘Wace closet privte tanc type 4 3 x Wace closet privite presure tink 43 u Waeer closet pubic dust valve 6 3 1 Waer closet pubic tank type 43 x Weer closet pubie pressure tak 4 3 xu Fixure not listed 1 om om Finure not listed 2% 1% Finure not listed 3002 Fixnre not listed So} aM Faxure not listed 6 4 2 Hoe bibb or sll-ock, public s% Hoe bibb or sill sock, private 304 ‘Waer supply notisted 1M ‘Waer supply notlisted 2 4 Waer supply notlsted 3K ‘Water supply not listed wo 1PLUMBING PPING SYSTEMS cent Note! For sudden or talon Fe tntracions, ate 1b “Lameter om tbe ominalpipe'ae real Seaxdyrd Tee ‘Diameteof Pipe, Tnches ‘Throush Side Outle: ie 3 ‘ AR a FOxlinaryEntrarce, [> “| Ns ; 2 dard Elbow ort of Sonera ratce : QB Medium 3rcep Eloy un of Tee reduced i Ga Long Sweep Elbow or franto'Sandaed Tee ot FIGURE C131 Resstance of valves and fitirgs to fow of fiuids. (Counesy Crane Co.)cs72 PING SYSTEMS along with aay restrictions on their use. The code alio ‘may stipulate accepted standards thet govern the manufacture of the compon:nts, their tolerances, and their installation. The piping materials to be discussed are all accepted for use in various natiozal, regionc, and mos local codes. Some materials may not be accept abe for use in specific local codes When a plumbing system is beng renovated or when circuristances require a unique design, it may b> necessary to request deviatior from the accepted Ist of materials in order to match existingpiping or to obtain special desigr characteristics, ‘Te review such requests, the authcrities require enough information to determine whether the intent of the applicable code provisions is achered to ir terms of safety ani suitability of the materials for the intended purpose. The design engineer's ultimately responsible for selecting and specifying the ‘mest suitable pipe, fittings, and jcinting methods for any project The following, characteristics of pipe and fittings are important considerations in making thet se- lection: 1. Corrosion resistance of the pipe and fittings. This is a measure of their ability to resist bth the intemal corrosive effects of the fluid likely to flow through the pipe and the effects of soils or ambient conditions on ts exterior. Corrosion can. be reduced or eliminated by the application of a suitable coatirg, encasement, lining, and cathodic protection, 2, Total installed cost, which includes the cost of the pipe and fittings, assemlly of tHe joints, handling, nd the cost of the support system for the >iping. 3, Physical strength of the pipe ard fittings, which is the ability to withstan¢ the internal pressure of the liquid and external physica damage that may cccur sither during installation or after being placed in service. Metallic Pipe and Fittings Cast-Iron (C2) Soil Pipe and Fittings. Cast-iron soil pipe is acceptable for any noapressure, aoncorrosive sanitary and stormwater drainage serviee. Three types of ripe arc manufactured: service lor standard) weight, extra heavy, and hutless. ‘Two types of pipe ends are available: hub and spigot, ard hubless. Three types of joits are used: caulked and compression gaskets used vith hub aad spigot pipe, ‘and compression couplirgs used only for hubless pipe. Hlub and spigot cast-iron soil pipe and fittings shall con‘orm to thefollowing :tan- dards: 7 1. ASTM A 4, Cest-Iren Soil Pips and Fittings 2, ASTM C 4, Rubber Gaskets for Cast-Iron Soil Pipe and Fittings 3. ASTM HSN, Nenprene Rubber Gaskets for Hub and Spigot Cas: Iron Soil ?ipe and Fittings 4, CISPI 301, Hubless Cast-Iron Soil Pipe and Fittings for Sanitary and Storm Drain, Waite, and Vent Piping Applications 5, CISPI 310, Patented Joint for Use in Connection with Hubless Cast-Iron Sani- lary System Acid-Resistart (AR) Cast-Iron Pipe and Fittings. This pipe material is usec for nonpressure drainage service for ccrrosive liquids whose corresion sotential istoo severe for Cl pipe. AR aipe is CI pipe containing between 14.25 and 15 percent silicon and small amounts of manganese, sulfur, and carbon. It is manufactiredPLUMBING *IPING SYS"EMS ceva only in extra-heavy grade. It is available with tvo types of pipe ends: hub and spigot, or hubless. The hub and spigo: ends can be joined by caulkirg. Hubless pipe is joined by the use of compressicn coupling. -Acid-resistant cast-iron pipe and fittings mast ccnform to tke following standards: 1. ASTM A 518, Corrosion Resistant High Silicon Iron Castings 2. ASTM A 861, High Silicon Iron Pige and Fittngs Ductile-Iron (D2) Pipe and Fittings. DI pipe 's suitable for any noncortosive plumbing service, Ductile iron pipe is fabricated of 3 cast-iron alloy in which graphite replaces the carbon that is present in cast-iron scil pipe. It s available for use as either a nonpressure gravity sewer pipe or a pressure pipe. Five pressure classes (@C) 150, 200, 250, 300 and 350 are available, Ductile iren pipe rating is also designeted by Special Thickness Classes: 50, 51, £2, 53, 54, £5 and 56, and gravity sewer pipe, Three types of joints are ued: mechenical, gasketed, and tanged. Duaile-iron pipe and fittings must conform to the following standards: 1. ANSI/AWWA C1SI/A21.51, Ductie-Iron Pipe, Centrifugally Cast in Métal Molis or Sand-Lined Molds, for Weter or Otter Liquids 2. ANSIVAWWA C115/A21.15, Flanged Ductile-Iron and Gray-Iron Pipe with ‘Threaded Flanges 3. ANSI/AWWA C111/A21.11, Rubber-Gasket Joints for Ductile-Iron and Gray- Iron Pressure Pipe and Fittings . 4, ANSI/AWWaA C110/421.10, Gray and Ductile Iron Fittings, 2' though 48 in for water and other liquids For other standaris refer to Table C1. in Chap. Cl. Steel ($7) Pipe. Steel pipe is commony used in vent systems, in drainage systems where auman waste is not discharged, for: indirect waste lites, for potable water ‘systems, and for fuel gas-piping. Steel pipe can te obtainec with threaded ends, plain ends, and.teveled ends. Four types of joists are used: screwec, grooved, flanged, and welded. Fitting materials commoniy ised are steel (for welded joints), malleable iron (either screwed or grooved joints), and cast iron (either screwed or flanged joints). Steel pipe for plumbing :ystems should be galvanized toretard cor- rosion. Sted pipe and fittings used for plunbing systems must conform to the Tollow-"~ ing standards: 1. ASTM A 53, Pipe, Steel, Black, and Hot-Dipped, Zinc Coated Welded and Seamless 2. ASME B16.1, Cast-Iron Pipe Flanges and Flanged Filtings, Classes 25, 125, 250, and 800 3. ASME BI6.3, \lalleable-Iron Threaded Fittings; Class 15) and 300 4, ASME BI6.5, Pipe Flanges and Flarged Fittinys 5, ASME B16.9, Wrought Steel Buttwelding Fittings 6. ASME B16.12, Cast-Iron Threaded Drainage Fittings Coppe: Water Tube. Copper tube is used for domestic water service. Copper tube is fabrizated of 95.9 percent copper with plain ends, in three types: K,L, and M. Each type has the same outside diameter, with K tube having the greatest wallca74 PIPING SYSTEMS thickness anc pressure rating, and M the least. Bach of the three types of copper tube is also available in either drawn (hard) or annealed (soft) forns. Fittings can be either wrought copper or cast bronze. Copper tube can be joined by fhred, soldered, or brazed joints. ‘Copper tube and fittings must conform to the following ASTM and ASME stan- dads: 1. ASTM B $8, Seamless Copper Water Tube 2. ASMEBI5.22, Wrought Copper and Copper Alloy Sdder Joint Fressure Fitings 3. ASME B16.18, Cast Copper Alloy Solder Joint Pressure Fittings 4, ASME B16.26, Cast Copper Alloy Fittings for Flared Copper Tubes Copper Tub, Type DWV (Drainage, Wasie, and Vent). It is a nonpressure, thin-Wall drainage pipe, primarily ised in residential buildings, and in commercial buildings for indirect waste lines or local branch lines vhere human waste is not discharged. I: is a.seamless tube, made from almost pare copper (99.9 percent), ‘and is available only in drawn form with plain ends. Joints can be either soldered or brazed. 4 Type DWY copper tube for drainage systems ‘must corformto the following stan- ar 1. ASTM B 306, Copper Draitiage Tube (DWY) | Sin | fen 2. ASME B1528, Wrought Copper and Wrought Coppe: Ale Soller Joint ran- age Fittings, DWV 3. ASME B16.23, Cast Copper Alloy Solder Joint Drainage Fitings, DWV Brass Pipe. It is genesally used in local branch ‘draliage nes vbere this aloy reiists specific corrosive drainage effiuent, in alterations to’match existing work ani asa pressure pipe for potable water in sizes larger than NPS 4'(DN 100), ‘where soldering or brazing is impractical. Brass pipe is manufactured from an alloy containing 8£ pereent copper and 15 percent zine with plain ends. Joints can be either screwed, solderec, or flanged. Brass pipe and fitings met conform to the folowing standard 1. ASTM B 43, Seamless Red Bress Pipe 2. ASME B16.24, Cast Copper Alloy Pipe Flanges anc-Flanged Fittings; Casses 150, 300, 400, 600, 909, 1500, ard 2500 3. ASME BI6.15, Cast Bronze: Threaded Fittings, Classes 125 anc'250 4, ASME B16.18, Cast Copper Alloy Solder-Joint Pressure Fittings Plastic Pipe and Fittings Gmeral. Phstic pipe s manufadured in a great variety of compositions, nany ofwhich are suitable for plumbingoystoms. Tao applicable code is ihe most impor- tant factor in selecting the type of plastic pipe for any soecific purpose. All plastic pipe, components, and jointing methods used in potatle weter systems must be ‘approved by the NSFI. Plastic pipe must be closely integrated with the selestion ofhangers and the entize pipe sugport system. ‘The advantages of plastic pipe include excellent resistance to a wide Variety ofPLUMBING PPING SYSTEMS C675 chemical and waste effluents, resistance to aggressivé soils, availability in long lengths, low resistance to fluid flow, ard low inital cost. Dsadvantages include poor structural stability (requiring addiiional supports), lower pressure ratings at elevatec temperatures, susceptibility of some type: of plastic: to paysical changes due to exposure t» sunlight, low resistance to solvents, and production of toxic {gases released upon combustion of som? types of slastics, Thre: designations are used to express pressure Jating and vall thickness: sched- ule (dimensions are outside-diameter-coatrolled, matching iroa pipe size} standard dimensional ratio (SDR) (a pressure rating only); and dimensional ratio (a pressure rating only using nonstandard dimensional ratios). Where the oressure rating is the prime consideration, the abbreviation PR is used. - Polyvinyl Chloride (PVC) Pipe. PVC is used for potable water and drainage systems. It is one of the most widely used of the plestic pipes. [t has a lov pressure and temperature riting and very poor resistance to solvents. PVC pipe and fittings must conform to the following standards: 1. ASTM D 1785, PVC Plastic Pipe, Schedules 40, 80, and 120 2. ASTM D 2241, PVC Pressure-Rated Pipe (SDE Series) 3, ASTM D 2466, PVC Plastic Pipe Fitings, Schedule 40 4, ASTM D 2467, Socket-Type PVC Plastic Pipe Fittings, Schedule 80 5. AST D 2665, PVC Drain, Waste aud Vent Pize and Fiuings Chlorinated Polyvinyl Chloride (CPVO Pipe. CPVC is used for potsble water and drainage systens, It has the same characteristics as those of PVC aad is used where a stronger piping system with higher precsire and temporature ratings io requirec. CPVC pipe ard fittings must conform to the followng standaids: 1. ASTM F 441, CVC Plastic Pipe, Schedules 40 and 80 2, ASTM D 2846, CPVC Plastic Hot and Cold Weter Distritution Systems 3. ASTM F 439, Socket-Type CPVC Plstic Pipe Fittings, Schedule 80 Polypronylene (PP) Pipe. This materal is widey used fo: chemical drainage piping systems. PP sipe and fittings are manufactured from flame-retardant material and are available in Schedule 40 or 80. Joining methods incude solvent cement joints, threaded joints, or mechanicaltype joint. (Only Schedule 8) can be threaded.) Polyethylene (PE) Pipe. tis widely used for undeiground fuel gas and foundation drainage piping. Itis joined by socket and butt hea: fusion. Refers to Chap. Dl of this handbook. Acrylon'trile-Butaliene Styrene (ABS) Pipe. ABSis widely wed as drainage pipe and is available in Schedules 40 and 80 vith plain cr socket eads. Joints are made by ithersolvent cenent or threaded connections. Orly Schedul> 80 can bethreaded, ABS pipe and fittings must conform t» ASTM Siandard D 2661, ABS Schedule 40 Plastic Drain, Waste, and Vent Pipe and Fittings.6.676 PIPING SYSTEMS Other Glass Pipe. This pipe is primatily used for gravity d-ainage of various corrosive Iquids. Glass pipe is iabricated ‘rom a low-expansior borosilicete glass heving a low alkali content. Glass pipe is joined by compressioa coupling: Vitrified Clay (VO) Pipe. This pipe is suitable for ase in underground gravity ctainage systems where resistance to a wide variety of corrosive efituent and eggres- sve soils is required. t is manufactured in standard and extra-strong grades With ub and spigot ends. Clay pipe is joined by mortar joints. MISCELLANEOUS FITTINGS Adapters are used to join two different pipe materials o piping with dissimilar joint ends. Most plumbing codes require the use of approver! adapters Dielectric fitings are used to connect dissimilar metallic pipes together toavoid gulvanio corosion that would quickly weaken the pipe at the joint. The prnciple of the joint is a gasket that isolatss one pipe from the other. 2 Unions ace fittings wsed to comect two fixed pipes, neither of which is cepable 0! being tured: The union consists of three interconnected pieces: two internally threaded enis and a centerpiece ‘hat draws the ends together wken rotated JOINTS A aoa A joint is a zonnection between one pipe and either another pige of a fitting, It must be ablé to withstend the greatest pressure capable of being exerted ugon it, Most plumbng codes refer to stendards that govern the methocs and materials used in forming joints. The selecton of the joining methods is detcrmined ky the ‘type of pipe and fittings used, the maximum pressure expected in the system, and the need for disassemby, Caulked Joints A caulked joint, illustrated in Fig. C132, is a rigid, nenpresourotype joint, This Jont consist: of @ rope of oakum or hemp that is packed into the annular space around the spigot end. For acid-resistant cast-iron pipe, hydrous magnesium alumi- num silicate, reinforcec with fiberglass, is used as a packing material instead of odvum. Molten lead 1 ia in depthis then poured into the anaular space on top of the rope. The lead is then driven (caulked) ferther into che joint. 11 use, the }emp oroakum svells when it absorbs vater and further increases the oint’s abil: ity to retist leaking Because caulked joints are labor-intensive, they have generally been repaced by either compression coupling or gasketed joints for most cast-iron joint applica- ticns where permitted ty code.PLUMBING PING SYSTEMS C677 pOOvE Iw vue Ee EROS Fateh Orne Beka RIS ARNULAR fPace, WHEN sPICoT END se Sine ptr seen ney iestte FIGURE C132 Caulked joint. Compression Couplings. ‘The comproscion coupling, illustrated in Fig. C13.3 is a rigid, nonpressue type of joint that can be easily disassembled. The coupling 2onsists of an inner elastomeric gasket and an outer metallic sleeve with an integral bolt used for tightening and compressing the gasket. Pips ——». Gasket Stainless steel ‘Sainess steel shiid ——e retaining clamp Tightoning bolts FIGURE C133. Compression coupling. . This joint is preferred for abovegrounc instellations because of its ease af assem: bly and strength. Underground, the metalic sleeve often fails after years of service due to corrosion by surrounding soil or fil. The stardard goveming the fébrication Of this type of jointis Cast Iron Soil Pips Institute (CISPD) Standard 316. Screwed Joints ‘The screwed joint, ilustrated in Fig. C14, is a rigil, pressure type joint that can be easily disassembed. Such a joint can >e used with any plain-end pipe that has the necessary wall strength and thicknessto have threads cut into it. The joint uses threads cn two pipeends (or on a pipe and a fitting) 10 draw thetwo piecestogether and form a leakproof seal. The threads used for pipe are known as American (tapered! pipe thread (APT). This type of ointis generally limited to NPS 3(DN 80)e678 PIPING SYSTEMS ‘Thre angle 60° ‘tom the 1001 Piteh cf the thread crest or top” of trea ty “tres "of tread Impetect iveads Effective threacs FIGURE C134 Scewed joint. Pipe and smaller in diemeter, because of the great effort required to turn ¢ pipe” of larger size in making a joint. . ._. Applicabie plumbing codes usually specity the type cf pipe that 2an be threaded. ‘Tre:standard governing the threads rae Roar Pipe Tareads, General Pirpose, ys mat a cae oy Soldered Jcints Asoldered joint, illustrated in Fig ‘C1357 is a rigid; pressure-type joint made with a filler metal called solder that,-when heatéd tovits melting point, is drawn by capillary action imto the annulaf space between the ipe and fitting. Whe the solder cools, it adheres to the walk of both pipe aad fiting, creating a joint that is suitable for iny installation for waich the piping itself is acceptatle, This type of joint is generally limited to pipe having diameters no larger than NPS 4 (DN 100) because of tke difficulty of applying heat evenly’to larger joints. Solder melts ata temperature of 900°F (482°C) or lover. When tsed in drainage Capillary space (exaggerated for istration) Copper water tube ‘Solder or brazing alloy FIGURE C135 Sadored joint.PLUMBING HIPING SYSTEMS C679 and vect systems, an alloy of tin-and lead or tin aad antimony is used. Solder and fiux used for potaole water systems must contain 10 lead. ‘The solder material should conform to ASTM Standard B 32, Solder Metal. Brazed Joints A brazed joint is similar to a soldered joint, except for the melting point of the filler metal, which is higher than 900°F (482°C). This joint $ used where hi pressure ratings are required than are sed for soler joints. Yarious compositions of the fller metal are available for varisus applications. Brazing metal should conform to AWS Standarc AS.8, Filler Metals for Brazing. Gasketed Joints : oh A gasketed joint, illustrated in Fig. C136, is a flexibié, presure-type jpiit using an elasomeric gasket under compression. The joint is capable of being easily disassernbled, It is well suited for both aboveground and unde-ground installations. VE Adjoining pip» Hut FIGURE C136 Gaskets joint. ‘Various manufacturers produce pipe ends and gasket configarations for different applicatons that are not compatible with one another. , Standards goveming the fabrication of gasketed ints are ASTM C 564, Rubber Gaskets for Cast on Soil Pipe and Fitings; ASTM D 3212, Joints for Drain and Sewer Flastic Pipes Using Flexible Elasomeric Seals; and CISPI HSN, Cast Iron Soil Pip:. Mecharical Joints ‘A mechanical jointis a pressure-type joist that can easily be disassembled and uses anuis anc bolts to draw together a pipe ad gland te compressa gasket aound thec.630 PPING SYSTEMS Not (Gland Pipe Pipe Keskes FIGURE (13.7 Mechanical joint. Pipe, forming a leakproof seal. A typical joint is illustrated in Fig C13.7. Many different kinds of proprietary meckanical joints are avaiable to achieve varying dogiees of flexibility and pressure rating. Grooved Joints A grooved joitt, illustrated in Fig. C13.8, is a pressure-type joint for metallic pipe that can be eaiily disassembled. The joint consists of an nner elastomeric gaccet ‘oli rea retallic coupling -Elistomeie ‘gasket SIGURE C188 Groove! joint and an outer split metalli: sleeve with an integral bolt wed for tightening. The ‘outer sleeve has extensiors at each end that fit into grooves cut in or rolled irto the ripe near tke ends to be joined. Roll grooves are used to form jonts when the Pipe wall is not thick enough to have ¢ groove cut in it. Different stylesare availatle to achieve varying degree: of fexibiity and prosoure rating, It is highly resistant to being pulled apart. Standards governing the fabrication of grooved joirts are AWWA C606, Grocved and Shouldered Joints; ASTM D 735, Gasket; and ASTM A 193, BolsPLUMBING PIPING SYSTEMS C681 Gasket Flange Flange Threads“ Fipe FIGURE CBS Flange joint Flanged Joints A flanged joint, ilustrated in Fig, C139, is a rigd, pressure-type join: that uses nuts and bolts through a raised projecton on the end of a pipe to draw the ends of the pipe together against a gasket, io form a leakproof joint. It can easily be disassembled. Flanges can be cast integrally, or they can be welded or screwed onto a plainend pipe. The face of the mating flanges cin be raised or flat. A variety of proprietary metheds of flange attachment to plain-end pipes are also available. Welded Joints ‘The welded joint, Iustrated in Fig, C1310, is a pressure-type joint most often used for steel pipe. Welding is accomplishec by bringing both pipe walls, at the joint, Prepared Files pipe end metal Butt weld For metal thickness of Sfesinch or less () Fitting or Pipe soupling a . aw Fille: » metal FIGURE C1310 (¢) But-weld joint. () Socket weld jont.csa2 FIPING SYSTEMS to the melting point and fusing them together wita the addition o: metal to allow fo: correct wall thickness and streagth. The necessary amount of heat for welding is produced by either a high-temperature flame or ani cectric arc formed between the welding electrode and the pipe. To properly butt-wed pipe, the pipe ends must be specially oropared depending on the pipe thicknes:, pipe metal compos tion, and welding method. Two types of joints are possibk: butt, illustrated in Fig. C13.10a, and socket, illustrated in Fig. C13.105. The foar method: of welding are gas tungsten arc weldiag (GTAW) or TIG welding; gas metal are welding (GMAW) 0: MIG welding, shielded metal arc welding (SMAW); and oxyfuel torch welding x Coupling shank Flared Joints Thread ‘This is a rigid, pressuretype joint used . - only with annealed (soft) copper tubing. Flare A flared joint, illustrated in Fig. C13.11, oupting net is nade by first placing aloose, threaded ‘coapling nut on one end of the pipe,then cod-forming that end with a maadrel that enlarges the pipe end to fit a mating Tube end on a thriaded coupling shank. The sciewed coupling nat and shank arethen tumed, drawing the pip: ends together to form a leakprocf seal. FIGURE CBA1 Pipingsystems JOINTS FOR PLASTIC PIPE Sdlvent Cement Joints ‘This is a rigid, pressuretype joint that is suitable for ary type of installation for which the piping itself is acceptable. A solvent cemert jomt, illustrated in Fig. C1B.12, is mate by sprezding a combination of solvent aad cementon the surlaces Socket Wped bead spas ODL LIZA Prastic sipo L area of solvent and cement application FIGURE C13.12 Sovent coment joint for platie pipe,FLUMBING PIPING SYSTEMS ce8g tobe joined. They react chemically, dissolving the strface of the pipe and the fitting thar it comes in contact with. The two componens are put together vhile wet. After drying, the tvo components are fused into a homogeneous mass, producing a leakproof joint, Solveat-weided joints can be used only wih specific matching types of plastic pip and fittings with plein and soccet ends. The Hllowing sandards govern the use of solvent cemen:, depending on the type-of pipe for which the cement will be used: ASTM D 2235, Solven: Cement for ABS Plastic Pipe and Fittings, and ASTM D 2564, Solvent Cements for PVC Plastic Piping Systems. Heat Fusion Joints » A heat fusion joint i a rigid, pressure-typejointthatis only suitable for thermoplestic pipe materials. Heat is used tomelt the plastic pipe surfaces and fuse them together into a homogeneous mass. Two types of jaints are available: but, ilustrated in Fig. C13.13a, and socke; illustrated in Fig, C:3.13b. Such a joint can be used only with Plastic Excess plastic Pipe mateial \ f Pigtail tb) FIGURE C1313. (2) Socket hea-fused joint for plastic pipe. (2) ‘Butt hea-fused joint lor plestic pive.cs84 PIPING Systems ‘two plain-end plastic pipes (butt joints) or ¢ plain-end pipe and a socket fiting with resistance wire insde the socket manufactured spscifically for this purpose. Fer butt heat fusion, the ends of the pipe are heated tothe melting point with an ouside souree of heat, usually a fiat, electrically heated plate. The heated plite is removed, and the two ends of the pipe are brought together, The socket joint is made by placing the plsin end of the pipe into the socket, connecting leads ‘rom wire embedded inside the fitting to a proprietary electrc source, and heating the socket to the melting point to fuse the pipe to the inside o° the socket, The manufac turer’s instructions mustbe followed carefully throughouteach phase of the process. Its accepted practice to have the mechanics making up the joint sertified by the manufacturer as being properly qualified in the correct >rocedures The standard governng the fatrication of heat-fused joints is ASTM D 2657, Heat Joining of Polyolefin Pipe and Fittings. For more details on nonmetallic piping systems, see Part D of this book. SANITARY DRAINAGE SYSTEM PIPING System Description The sanitary drainage system conveys waterborne’ sffiuent dicharged from plunbing fixtures and cther equipment, to an, approved point of isposal. The ly treated sahitary system receives all liquid waste except stormwater or unaceept process or chemical drainage. 2 ‘System Components Maor components of the sanitary drainage system ‘are pipe and fittings, joints, valves, traps, dleanouts, drains, interceptors, sewage ejecors, and simp pumps Nomenclature Approved: Accepted for the intended purpose, a8 an appropriat> design-or for ‘nstallationinto 2 plumbing systen by a responsible code official or other agency ‘exerting jurisdiction for a specifi: project. Backwater valve: A commonly used term for.a type of check valve used in. a crainage system. - Branch: 4, horizontal run of pipe not considered a house drain or stack. Branch inteval: The distance measured along the stact, within which horizontal crainage branches are connected to a drain stack. Thi: distance is usually cne- Story heigh’, but never less than § ft, 0 in (2.4 m). Building drain: The west horizontal part of the drainage piping system; con- sidered the >rincipal pipe couveying sanitary effuent bygravity to & point outsde te building. Building sever: The continuation of the building drain from a point outide ‘he building wall fo the actual connection to an adequate end approved pcint cf disposal, such as a public sewer or private sewage cisposal system.6 PLUMBING MPING SYSTEMS c.6as Buitling wrap: A wap installed on the house sewer to prevent the circulation of sewer gas between the building s:wer and tre building drain. Chemical waste: Any substance that may cause harm to the sanitary piping system, treatment facility, or enviroamen: without being treated or neutralized prio: to discharge into the sanitary drainage system. Cleanout: A gas-tight, water-tight pipe fitting with a removable plug that is ‘usec to obtain cccess to the inside of « drainage pipe for cleaningor méintenance. Combined dratsage system: A draixage system that combines sanitery effluent and stormwate: runoff in a single piped system. Fitting: A device used to connect one or more pipes ind/or to change the direction of a straight run of pipe, Flocr drain: a plumbing fixture that removes liquid effluent from the surface of floors and ocher areas. Howse drain: & commonly used tem for buiiding drain. Howe sewer: A commonly used term for buiding sewer we Howse trap: commonly used tern for building trap. Indivect waste: Any waste pipe not connected firectly into the drainage system that discharges through an air gap into a fixture, intercepoor, trap, or drain. Interceptor: A device that séparates, retains, and allows removal of specific harmful material suspended in the waste siream, while permitting the remaining acceptable liquid effiuent to be discharged inte the draincge system. Invet: ‘The'elevation of the inside bottom of a drainage pipe. ‘ Leader:'" A vertical or horizontal lite conveying stormweter. : Offset: Any change ‘indirection of stack from vertical.’ 1 aye Pitch; The distance iat one end of « pipes lover thai the other end, éxpressed. as apercent of the total length of rin or as.a dimension, in, inches or feet per foot of rua": ‘Plunbing fictues. Any approved. receptacle o: device specifically designed to receive human ar other waterborne vaste and cischarge that waste directly into the sanitary drainage system, often vith the adiition of water. Runout: A commonly used term for the first section of horizontal drainage piping from a stack at its lowest level. Stop: A commonly used term for vitch. Soil line: Any pipe that conveys human waste. ‘Stack: A vertical drainage line, usually more than three floors in height. Trap: A device that maintains a water seal, preventing “he passage of sewer ‘20s, vermin, ait, and odors originating from inside the dminage system, while permitting the inrestricted passage of liquid waste into the drainage system. Waste line: A drainage pipe conveying liquid waste with no solids. Major System Components Cleanoxts. Codes mandate that cleanouts genenilly be provided at the base of stacks before the pipe changes directior from vertical to horizontal, at changes inC686: PIPING SYSTEMS seh ri Ferrule @ GURE C1314 (¢) Typeal cleanout, (0) Cleancut components. borizontal pipe direction greater than 45°, and along horizontal rans of pipeevery 5) ft.(15.m) Typical clanouts ar illustrated in Fig. Ci3.14¢ and cleanout compo. nents in Fig C13.14b. . Hloor Drairs. A floor drain is « receptacle used to remove liquid effluent from building interior foor areas and cther locations. A typcal floor dain is illustrated it Fig. C1315. It provides a receptacie for spills, waihdown; ard effluent to be collected and routed directly into the senitary drainage piping system. Code provis- ens do not specify where a drain should be jocated, Hewever, most codes rezulate te minimun seal requirements for drain traps, the minimum open area of grates and strainer, and the mandatory inclusion of certain irdividual components (such ‘as removable secondary strainers or seciment buckets) for drains in some locations. A standard commonly cited in the selection of floor drains is ANSI A112.21-1. Foor Drain, Dreins consist of tke following components: 1. Drain body : * Grate ‘Secondary, Sediment L 4, svainer bicker Having “ fl oie eat ie tinge - Flange: Body outer Dutiet outer FIGURE C135 Typical for drains,PLUMBING PPING SYSTEMS C687 2. Grates located at the top of a drain permit liquid effluent to enter the drain body while excluding larger solids and foreign matter. Grates are classified as follows: @, Light duty: fot traffic only . 3. Medium duty: live wheel loads up to 2000 Ih (907.2 kg) Heavy duty: live wheel loads up to 5000 Ib (2268 kg) d, Extra-heavy duty: live wheel load; 5000 Ib (2268 kg) 01 more 3. A secondary strainer may be installei below the grate in ¢ drain tha: does not havea sedimen: trap. 4. A sediment trep (or bucket) is a -emovable device indde the d:ain body that may be imtalied to trap and retain small solids that pass th-ough the grate 5, A flashing ring or clamp is a device vsed to sectre flashing directly te the body of the drain Interceptors. Plumbing codes require that any su2stance harmful to the building drainage systém, tae public sower, or the municipal sewage treatment process be prevented from beng discharged into the public sever system. Among such materi: ‘als are zrease, flammable liquids, sand, or other substances objectionable to the local authorities. Traps, A fixture trap, illustrated in Fig, C13.16, s a U-shaped section of pipe of, the necessary deph to retain sufficient liquid required by code. All fixtures and cquipmnt directly connected to the sanitary drainage sysiem are required to have traps. Outlet - Drown Weir Dip rap Cleanout FIGURE C1316 Typeal fixture tmp. In general, traps must (1) be self-cleaning, (2) provide a iquid seal of at least 2 in wih larger seals where required, '3) canforn to local code requirements in terms cf minimum size, (4) provide an accessible cleanout, znd (5) be capable ofcess TIPING SYSTEMS rapidly drairing a fixture. All trap: must be vented ir an approved manner, ex- cept for specific condiions waived by lecal code requirements or authorities. ‘Traps that are prohibited by cole in- clide traps requiring moving pats to ‘maintain theseal; full Sype traps, illus- trated in Fig.C13.17; crown ventedtraps, illustrated in Fig. C13.18; and drum traps, illustrated in Fig C1319, Drum traps may be permittee by some 2odes foc use on special-use sinks, such as in laboratories. ’ ‘The branch drainage line extending from the trap to the vent is called the trap ann and is ilustrated in Fig. C13.2). The maximum lengh of the trap arm is siown in Table C12, FIGURE C1317 S trap. Sanitary System Design ‘The design o! the gravity drainage piping system is strict regulated by the applica- ble plumbing code. All'codes inchde charts, similar to those presented here that permit the design engineer to properly size all horizontal and vertcal pipes based on the accumulated drainage fixture unit discharge and slope of the pipe. Vent Crown if less than two pipe éiameters trap is crown vented FIGURE 013.18 Crown vented trap. ‘The pitch of the drainage system must provide sufficient veloci:y to produce a “scouring action” that will convey all solids along with the ligud stream. The recommended minimum velocity for ordinary sewage is 2 ft/s (0.6 m/s) to prevent the settlemert of solids out of the effiuent stream. When grease is in suspersion, the velocity should be at least 4 fis (1.2 m/s) Accepted practice fo: low-rise buildings with relatively uniform discharge is to size horizontal dréinage lines to fow half full under average design conditions. High-rise buidings produce higher velocities and turbuleace in building drains that cas fill portions of the pioing system as much as three-qucrters full, vith completely full pipes expected for stort distances at stack runouts. This is eccepiable, providing thet pipe size ultimately allowed for half-full pipes.PLUMBING PIPING SYS"EMS C688 ‘The followingis a simplified methoc of sizirg the drainage piping system: 41. Esteblish the Iccation, size, and inver: ‘of the point of ultimate disposal 0: sanitary effluent, Determine whethe: sump oF ejector systems will be re quired and locate them. 2, Locate and lay out drainage branch lines, stacks, aad the house sewer. 3. Start with the individual device o: fixture at the farthest and most re- mote point of the system or branch, for which the code specifies two FIGURE C1319 Drum trap. drainage values, The firsts the drain- agefixture uni: (DFU) value that will be ased to size the drainage piping : system. The sevondis a minimum size of tie trap, which is the minimun individua. ‘branch pipe size. For the minimum: trap size, refer to Table C131 for typica values. Ifa fixure or device is not Isted, either use the unlisted value based or the size of the discharge or ask the local code official for the acceped value. INSIDE WALL OF Pie DISTANCE FROM /\ _ "TRAP TO VENT - AIGURE CB20 Trap am. i4 4, When more than one fixture or waste line discharges‘ ino a horizental branch pipe, the size of the horizontal draizage line is determined by both the pitch of the drainage Ine and the accumulated total rumber of DFUs discharging into it. Refer to Table C13.3 for sizes of branch ‘ines and for the size of building drains and sewers. TABLE C132. Maximum Length of Trap Arm* Diameter of trap arm, in Distance from trapto vent 1% 3 ft 6in 1% St 2 Bf 3 10f 4 wt For use with Fig. C1320, Lia = 254 mm +1 f= 030manid 108 1 ‘paniunad siosor soem ns Wem 240u 10M t ‘ponrutiod stop ‘yen oa) wen Sucu! ON } “panaod soso sem ON ooaz 00's si SL ones oosiz bas . oe tr oz 0057 006 wos ose OF xa 00%'T 009 cose: oe g dot au ose + C06 096, or sis au coz cor os oe StS ose au 06 os oe ot ory 405 au toe vw 18% Joe of Ie au 6 a 0 a sie 9 du 9 % or 9 ost dv Ku z ° + € oe cn oe. Koos — yess) BEY UF wpueiq NG SAN RreeL ToL, 80] 10 aaniay ——__—suorej0 ——_rewozoy was adi Or oon oon vow puis ou aay sype1s wor) soyouesG sow 9218 ‘usp Supping pue wresp Suipr ng Suidig oFeuresg Argues 40} spe07 UN aMxLY aFeuNIG aqISSIUIV WHEW E'ELD TTAWL c.690PLUMBING HIPING SYSTEMS c.691 TABLE C134 _Aprroximate Discharge Kaes and Velocities in Sling Drains (Flowing Half Full (Discharge rate and veloc) Pipe size Yuin/ft slope %ein/fislope Yeinlft slope ——_ Ye nt slope Disch. Vel. Disch, Vel, isch. Vel, — Disch., NPS DN spn fob gpm fpr gpm ps gpm, ™ 2 340178 we 40 391 142 5532.01 2 50 a2 1m 119 243 mh 6s os 1m 153 19 ©6282 3 80 16 19 28 25 351 319 4 100 1.36, 378 193, 334 273 755 3.86 5 ps 138 683° 223966 «316 = 3704.47 6 150 178 110 252 187.0 457 2220 5.04 8 200 217 200 307 MOO 4344800613 10. 250 252 436.0 356 516.0 504 8720 TAZ R300" 25310 4oL 80, S6T 14130, B02 * Hal full means ited toa depts equal to one-half ofthe inside diametes. 4 Conputed from tre Manning formule for haffull pipe, n= 0.015, Por one-quarter ful: Multiply discharge by 024, ‘or full: Maltipy discarge by 2.00.. ‘Maltply velocity by 0.70: Multigly velocity by 1.00. For three-quarters ‘ul: Multiply discharge by 182. or smoother ripe: Maltply discharge ‘Maliply velocity by 1.5. and veloty by 0.015 land divce by n value ‘of smoother pipe, $1 gm = 0075 Us es $1 fps = 0305 mi. 5, To determine tie size of ahorizontal drainage lite based ox flow given in gellons per minute rather than DFUs, refer to Teble C13.4. Use she appropriate pitch and /clocity conbinations necessary © select a sze. When there isa cembination flowof both DFUs and gallons per miaute into a horizontal line orstack, generally accepted practice is to assign two DFUs for eact 1 gpm (0.075 l/s) to alow sizing based on DFUs. ‘Thesize of a stack is governed by the total DFU discharge into it and its height. Refer to Table C133, using the appicable column and the total DFUs for the stack to find the stack size. To aze a stack based solely on gallons per minute, refer to Table C13.5. Two generally accepted recommendatiors regarding the maximum cross-sectional area which may be occupied with water flowing down a stack are % and 7/24, depending on the code used and the requirements of the cal authority having jurisdiction, Separate columias are provided fo: each of these two values. Ac. cepted practice is to use the ¥full criterion, which clossly approximates the allowable flow from a horizontal pipe flowing fill at quarter-in (6.3 mm) pitch. 8. If astack should offset more than 45° from the vertical, the horizontal offset portion of the stack must be sized es a house drain. If the offset size is larger than that portion of the stack higher than the offset, the larger size must be carried down from the offset to the lnwest level The portion of the s:ack above the offset may remain unchanged. 6.cz PRING SYSTEMS TABLE C13.5 Drainage Stack Capacity, gpm* Size NPS DN tall % full mm 2 5 6s 1% 40 ul 105 2 50 115 26 DA 65 318 4 3 80 324 672 4 100 ai 43 5 25 202 361 6 150 336 3 8 200 708 ns * For SI units gpm = 075 Us. 9. The purpose in differentiating between branch intervals and the actual nunber ofhorizontil soil or weste branchlines entering the stact is toprevent overloading ihe stack in a short distance. Many codes limit the number of D¥Us allowed in a branch irterval Flow conditions in the offset portion ofa stack oreate severe turbulence. Because of the resultizg pneumatic effects, all branch connections that normally would be ‘Upper section of drainage Vent sack—ol | Soll or waste stack 90° hoizontal offset B2ta is20 ool ZF e20 FIGURE C1321. Suds pressure areas. -PLUMBINGPIPING SYSTEMS C698 made at the level of the offset should be carried down 10 pipe diameters of the stack below the level of the offset. ‘Suds Pressure Areas. Appliances an¢ fixtures normally using detergeats, such as itches sinks, batatubs, showers, dishwashers, an¢ clothes weshers, could discharge a large quantity of detergents into the drainage system, During flow through the Grainaze piping, turbulence causes larg amounts 3f suds to te generated. The sud: accumilate in the lower portions of th> drainage system and can remain there for considerable time. When additional iquids flowinto these tections ofthe system. the suds are displaced and will follow the path of least resistance. Enough suds pressure can build up to foree the suds through a fixture trap. Suds pressure arcas ‘exist in the following parts of the drainage systen, as illustrated in Fig. C1321: 1. Foran upper-level stack offset serving fixtures on two or more floors above the offset, there are two suds areas. The frst area, C13.214, extends 40 pipe diameters of te stack upward from the base of the offset. The second, C13.2:b, extends 10 pipe diameters horizontally downstream fron the point of change in direction. 2, For an upper-level stack offset turring from Lorizontal back to vertical, there * is ote area, C.3.21g extending 40 pipe diameters of the stack upsteam from the fitting charging direction from horizontal io vertical. “ 43. Inthe horizontal runout froma stack wien the pe changes direction horizontally with a fitting greater than 45°, there are two aeas. The frst, C13.2:d, extends 40 pipe diameters of the horizontal pipe upstream from the change ia direction. ‘The second, C13.216, is 10 pipe diameters downstream. TABLE C136 Suds Pressure Area—Disance Determination For use with Fig. C13.21. “ Nominal pige size’ DN. 40 diameters* 10 diameters* We 40 Sft0im 2 ft 6 in 2 50 70 2106 in wm 65 shon 2M 0 in 3 80 «itn 2A 6 in 4 19013 fOin 5 £6 in 3 BS 17 ft0in aft in 6 150 200 =~" FH 0in 1 = 0308 m ‘When suds pressure areas are anticipated, no pipe shall connect to any of the ‘areas indicated as C1321a to Cl3.2le. Refer to “able C13. for actual distances based on pipe size SANITARY VENT SYSTEM PIPING System Description ‘The sanitary vent system is a network of pipes directly connected to the sanitary drainage piping system for the puzpose of limiting air pressuce fluctuations within the santary drainage piping to 1 in (£2.54 em) of water column.C494 PPING SYSTEMS ‘There are two primary reasons why the vent system is an integral and necessary adunct to any drainage piping network: 1. It prevents the loss of fixture trap ceals. 2, It permits the smooth flow of water in the drainage system Other lesser problems will be prevented if is excessive, sich as ir presstre in the érainage system 1. Unsightly movement of water levels in water closet bowls 2. The possibility of sewer gases discharging through a ixture tra 3. Noise in tke drainage system dve to the gurgling of vater System Corrponents ‘The components of the vent system consist of pipes directly coiected to the t require terminatior of vents through the roof. Code Considerations Mest probleres occurring in the drainage system not resulting from. blockagevare catsed by fluctuations in air pressure. These problems can be eithe: eliminated or Tecuced to a level where: they-are-no longer objectiomble, by-designing a vent system that linits these variations in the drainage pipirg network toa genecally accepted figure of 1 ix (+2.5:em) of water column. This basic cesign criterion has been used to determine vent sizing and allowabl: lengths ‘hat appear in madern codes. Nomenclature The following definitions are preseated to prevent any cifferences in terminology between this handbook and any local, regional, and national codes Branch vert: A brarch vent isa vent that connects one o: more individual or 2ommon vents to a vent stack ora stack vent. Circuit vert: A circuit vent is abranch veat that serves wo or more trapsand extends from a connection to a crainage line, in front of the last ixture conrec- sion, to a vent stack. Common vent: A single vent line serving two fixtures, Continuou vent: A vertical veat that is a continuation of the vaste line from a fixture to which it is connected Developed'engik: The total length ofa vert pipe, measured along the centeline of that pipe, from point to point. Fixture battery: Any zroup of two or more fixtures tha: dischargeinto a common sorizontal waste or seil branch.PLUMBING PIPING SYSTEMS C695, Individual ver: A vent that connects directly to only ore fixture and extends to ether a branch vent or a vent steck. Loop vent: A branch vent that serves two or more traps and extends from a point in front of the last fixture conzection to a stack vont, Maiz vent: Tae principal vent of a building. remaining undiminisied in from the connection with the drainege system to its termnal. Relief vent: An auxiliary vent that onnects tte vent stack to the soil or waste stack in-multistory buildings; used to equaliz> pressure between -hem. This connection will occur at offsets and zt set vertical interval: determined by code. event: Another name for an individual vent Stack vent: The extension of a soil or waste stack above ihe highest horizontal drainage connection to that stack. It is also the name of a method of venting using the stack as a branch vent comnection. Sud veruing: A method of venting in which there is a sids préssure zone. Trap arm: That portion of the drainpipe between the trap and the vent. Vem extension: The height of the vent above the roof al its terminal. Veni header: A single pipe at the highest level of a building conrecting the top of vent stacks in order to penetiate the roof only once. . Veni terminal: ‘The open-air location where the end of the vent stack is placed, generally above the roof, Vent stack: A vertical pipe extending one or nore stories and terminating in the outside air. Wetvent: A vent line that may algo sérve a3. drainpipa ~ ‘System Design ” . Genera! Vent System Design Considerations. Ditlerencesin pressure within drain- age pipng are caused by the flow of water. Whea water is lowing under desi conditions in a herjzontal drain (apprcximately cne-half ful), the ait above the liquid will be forced into movement by he friction between tre flowing water and the air. In a stack, the water flows arcund the perimeter o! the pipe, leaving a ceatral 2ore of air (except when overloided), ‘The following principles govern the design of tie vent system: : 1, Whea design flow is exceeded, the pipes ere completely filled with water. This compresses the air ahead of, and creates a vacuum behind, the solid frort of water. 2. The air movingin a Vent pipe has friction losses smilar to those of flowing water. ‘This is why the longer the pipe, the larger the size. 3. The amount of air displaced is proportional to the amount of water flowing in the drainage pige. The flow is determiaed by using drainage fixture unis (DFUs). 4. The size of a vent stack should be a ninimum o° one-half the size of a drainage stack The size of a branch vent shoald be a minimum of one-half the size of the branch drainage line it serves. 5. Ina plumbing code, where the neadirg for soil cr waste sizz refers to stack size, it should also te used for horizontal branch soil and waste stacks. Since the ventiag requirenents for a stack are more severe than those for a rorizontal drainage line, there is a small safety ‘actor.c.0968 PING SYSTEMS 6. All fixture vents must rise above the flood level of the fixture served so at not (o act as a waste line in the event the drain line beecmes blocksd, Daveloped Length Measurement, The developed length of an indvidual or com- ‘men vent is measured from its point of connection with the fixture trap arm to where it conrects with the branch vent or vent stack. The developed length of a brench vent is taken frem the farthest connection with a waste branch from the point being sized. The developed leagth of a vent stack istaken from its connection with the soil or waste stick to its tetminal above the roof. Siting of Vers Stacks, Vent Branches, and Fixiure Vents. Plumbing codes cortain the information necessary to size a vent system. A typical vent sizing chart is presented in Table C13”. To use Tadle C13.7, there are three itemis that musi, be knows: (1) the total DFU count 0: the soil or waste line associated witi the vent being sized, (2) the developed length of the vent being sized, and (3) the size of the soil ar waste branch orstack. Pe tes ‘ TABLE C13.7 Size and Length of Verte " CON scot” Draige Tae soil or fixtue » ef tes waste 9 unity Fee DN stack, NPI* conneted 1% 1% 2% 3 4S. 40. 1% t 2 L so 2 2 _ 6 Be a“ “ 3 we, 80 3° x 3 4 200° 100 4 200° ‘900° 4 500° 700 - 3 200 350 1s 5 SOQ? 2 300 5 1100 20 50 200. 704 6 350° P28 50" 200° 400° 1300 150 6 20 15 30 125% "30 “1100 ‘ 960 24 "100. 2 1000 6 1900 2 70 2% 700 8 600 50 15¢ 500 1300 200 8 1400 40 10¢ 400 1200 8 200 30 8-350 1100 8 3600 25 6 250 300 1000 75 425 1000 2500 30 100 500 3800 3080350‘PLUMBING FIPING SYSTEMS: . c.637 Having calculaied these items, entei the table with the most severe condition of soil sipe size cr DFUs. Then read horizontally across until you ccme to the figure that meets or exceeds the developed length that you calculated. Read up to find the correct size of the vent. Use the following as a guide to sizing: 1. For vent stacks, use the total DFU load for ‘he drainage stack and the full developed length of the vent to find ihe size, Vent stacks nust be undiminished in size for their entire length. 2, For branch vents, use the longest developed length from the point where the size 's being de:ermined to the farthest connecion to the waste line 3, For individual fxture vent size, refer to Table 213.1, 4, For building trap vents’ and fresh-ai: inlets; the size should be a mnimum of one-talf the siz of the building drain. | Vent Terminals, ‘Tae vent pipe passing wrought the roof must remain cpen under all circumstances. The two conditions that would cause the exposed pipe to become blocked are frost dosure and snow closure. Local codes and authorities will specify ‘or requre the min:mum extension of the vent pipeto avoid clostire by accumulated. snow on a roof. In the absence of specific code requirements, the followng can be used as a guide to locating vent extensions. The vent extension shall rot be located under, or withn 10 ft 0 ia (3.0 m) of any window; door, or ventilating opening unless it is 2 ft Gin (0.6 m) above that opening. If the terninal is though a buiding wall, it shall be located ¢ minimum of 1Oft 0 ix (3.0 m) from the property line, ¢ minimum. pf 10 ft.0,in (3.0.m) above grade, and rot under ny overhang Expérience has showa.that an NPS:4 (DN 100; pipe will >revent frest closure, Relief Vents. _ Sol or waste stacks with'no offsets, in buildings having more than. 10 brarch jaterya’s, shall be provided with a relef vent at each tenti interval, studrting at thé (03 floor. Offsets inthe drainage stacks may also be required, to have relief vents. a There areseveral acceptable configiztions allowed by Various codes. in deieral, the lower end of the relief vent shall ccnnect to the soil or waste stack below the horizontal branch serving the floor required to have the relief vent, The upper end of the relief vent shall connect to the vent stack no less than 3 ft (1 m) above that same fivor level. The size shall be equal to that of the ven: stack to which it is connected or the drainage stack, whichever is smallér. Circuitanid Loop Vents. ‘These venting schemes are intended to prove a more economical meansof venting than the individual vent. Its allowed only for venting of floormounted fixtures such as water closets, shower stalls, and floor drains and may net be accertable in all code jursdictions. Circuit veating is illtstrated in Fig. C1322. ‘Circiit venting requires a uniformly sized drainage line vith at leat two, but not exceeding eight, fixtures connectedin a battery arrangement. The drcuit vent connec to the vert stack from the horzzontel drain line from: a point between the two mest remote ixtures. In addition to the circuit vent, a rief vent i required to be connected t> the horizontal drain line at the end of ‘he battery, or every ight fotures. Thesizes of each shall be ane-half the size of th: horizontal drainage line or she full size of the vent stack, whichever is smaller. Loop venting is the same as circuit venting ezcept for the connection of the6.698 PIPING SYSTEMS T 4 Vent stack e bes A oT 4 stack oe . bey O04 4 KIGURE CB.22 Detail of circuit veat, . branch vent to the building system, The loop vent “lospi” back :o the'stack vesit indtead of the vent stack. This is illustrated in Fig. C13.23. " oe Wet Vents. A wet vent is a comsinéd vent-drain line that receivis drainage from fictures in iddition to serving at a vent pipe. Wet vents are primarily wed in residential-type projects and are permitted only in a linited numter of codes, Wet vents shall conform to the guidelnes provided in each respective code. ‘Showers Top toor FIGURE (1323, Detail of loop vent. -PLUMBING PPING SYSTEMS ces9 ‘Suds Relief Vents, Suds pressure zones are illustrated in Fig. C13.21. If adrainage connection to these zones is made, a relief vent must be installed from the base of the suds pressure zone of the drainage stack to a ronpressure zone. Typical suds pressure relief vent sizes are shown in Table C13. TABLE 13.6 Suds *ressure Reler Ve Drdin stack size Relief rent si NES DN NPS DN 40 2 50 50 2 “50 6 2 50 80 2 50 100 3 50 Q 1s 4 100 150 5 135 7 “ oe 200 Bovis 1507 2 ‘Simp and Ejector Vents.’ ‘AN codes require the venting of ejector pits, since they are gasketed and airtight. Many codes may also recuire the venting of sump pits. ‘The vent pipe size is determined from the gpm discharge of the pump and the developed length from the pit to its comection with the building vent system or vent terminal. Table C139 shows a typical method of sizing such vents. TABLE C13.9. Size and Length* of Sump Vents capacity of ‘Maximum equivalent length of vent, ft sump pump je 4! ve size‘of vent, NPS) 2 Spm lls 10 2 NL NL ONE ONL ONL 0 1 9 NE- NL ONE ONE, ca 75 200° NE ONL NLS 80 420 0380 NEN 100 2-7, 250. NL. NE 330 wow to 3 NE 200 NP. 0 6 0 NL 250 NP 10 360 12 NL 300 NE ne as 380 400 NP NP 108 a 210 500 NPUNP) NP oH 130 Developed lengh plus an appropriate allovance ior effets of entrane losses and ‘ction ‘duet tings changeindirestion, end changes indiameter.Suszested allowaaces may be obtained from NBS Monograpl 31 or other asceptatle souxes. Av allowince of 3 percent ofthe developed length may be assumed ifa more precise value is not available “No limit, actual values greater than $00 fc (132.5 a). 1 Liss than 10 fe 13 m). |§ Nat permitted. 0304 m,700 PIPING SYSTEMS TABLE C13.10_ Nominal Pipe Cros- Sectional Area, Schedule 40 Pipe NPS DN int % 40 1.767 2 50 3.1416 ers ro 4.908 3 80, 7.068 4 100 12.566 5 125 lo.635 - 6 150 28.274 Vent Headurs When several vent stacks are combined into a common header at the highest level to penetrate thr roof only once, various codes require that the ‘combined vent pipe be sized using the combined DFUs of all the connecte¢ vents ‘and the single longest developed length of all the vent stacks being combined. Other codes require that the comsined vent stacks have a minimum cross se:tional area of all the separate vent pipes being combined. Table C13.10 lists the cross- sectional arza for Schedule. 40 steel pipe: in square incies. : INTERIOR STORMWATER DRAINAGE SYSTEM PIPING Parpose ‘The purpose of the stormwater drainage system is tc collect stormwater runoff from building roofs and ancillary areas exposed to the weather, and to convey the runoff to aa approved point of disposal. Ancillary areas of a building include accaways, walkways, canopies, and balconies, Components Major components of tie stormweter drainage system are dreins and the pipeitselt, rain manvfacturers make various types of roof and specialty drains as vell as accessories sed for the installation of drains in all areas of a project. Any pipe material suitable for the sanitary drainage system is alto acceptable for use in the stormwater drainage system. Nomenclature Areaway. An enclosed excavated area below grede level end open 1 the weather. Canopy: A smell soof protecing a window or entiance. -PLUMBING PIPING SYST2MS 701 Conductor: Stormwater piping inside a building. ‘Design point: The specific point in tae piping retwork where pipe size is calcu- lated Downspout: A vertical pipe attached to gutters installed on the outside of a building Draix: A recestacle for the collection and removal of stormwater from strfaces exposed to the weather into the stornwater driinage piping network. Durcion: A commonly used term {or time of concentration, Flow rate: The mezsurement of a volume of water over time, sues as cubic feet 2er second : Frequency: Te estimated number cf years that elapse between the recurrence of storms with a specific intensity. . Gurr: An open horizontal channe. used to collect stormwater, usually made of sheet metal or wood, attached to the lowest point of a pitched roof. Imperviousness factor: A number iadicating the percent of rainfall available” as runoff and not absorbed into the ground, absorbed by plants, loft a: Pua, or lost to evaporation during the rainstorm, expressed as 2 decimal. Inlet time: A frequently used term ‘or overlard flow time. Intersity:' The rate at which rain fals for design purposes, measurec in inches per tour. Leader: A vestical pipe carrying stormwvate: either imide. or oxide ‘the building. set Overtow: A’ postive’ and feil-safe bitet for removal of stormwater that has reached a predetermined height above the roo! level. Overland flow ime: The.time rainvater takes to travelon the, ground from the farthest point of an outside area to a drain, measured in minutes, Pipitg neovork: , The entire’ stormwater drainage system: including all’ drains ‘and pipe, to the point of cisposal. Rate of rainfall: A commonly ised teim for intensity. Renin period: A commonly used term for frequency. . Runoff: "The actual dow rate ofstornwater discharging int the piping network. Scupner: A penetration through a parapet above the 10of level serving as an orerfiow. Sidevall area: Vertical surfaces that contribute -unoff to the stormwater system. Stormwater: Liquid effluent resultirg from any form of precipitatioa, such as rain, snow, hail’ or sleet. Time of concentration: - The length of time a rainstorm will persist ‘or design purposes, usually calculated by adding the overlend flow time to the time in pipe. Time in pipe: The length of time stormwater will take to reach one design point from another design point while inside the piping network. Tribwary area: ‘The entiro area contributing runoff inte any portion of the piping network,02 PPING SYSTEMS General System Criteria Goneral Information. The designof the stormwater dreinage system require: that the following informaticn be obtained to establish design criteria: 1. Local climatic conditons 2. Local building end plumbing code restrictions 3. Building we 4, Building construction and pitch of root 5, Location, size, depth, type, and availability of the ultimate point of disposal 6. Total size of roof and ancillary tributary areas 7. Method of connecting to publicsewers 8 Allowable methods of disposal and permits required if public sewers are not available . Client standards and preferences Désign Storm. The stormwater sjstem is designed to remove the maximum ex- pested runoffin a given >eriod. The ability to calculate the flow rate is complicated bby not being able to acurately p:edict many of the factors affecting the a:tual arrount of runoff resulting from any given storm. To calaulate the estimated maxi- mum runoff, an artificial “design siorm” must be created. This storm will serve as a Smolation capable of predicting runoff volume accurately enough to provide a basis for the design of the piping retwork. ‘The desigr storm is based on actual rainfall records and is plotted in convenient fomm by the National Oceanic and Atmospheric Administration/National Westher Service (NOAA). Special consideration should be given to the degree of protection provided for the building end its coments by the stormwater drainage system. This depencs on the importanze of the facility or the space below the rool, the importance of unnterrupted service, and the vaue of the equipmen: or matexal installed or stared. Since a severe thunderstorm or hurricane may produce rainfall rates lerger then those that the code has been based on, the prope-ty value may require ine design engineer to select a rainfall rate greater than the minimum provided for by the code. Roof Overffows. A fail-safe method must be provided to immediately renove exeess runoff from a roof before che water level rises to a point where danage could result. The most commonly used method employs scuppers that will sllow the excess runoff to be Jischarged directly off the roof and down the side o’ the bulding. Codes often stiaulate the aeight that scuippers saall be installed above the roof level anc the size of the opening, One disadvantagr of scuppers is that vave action may allow water to spill out o: the scuppers before the design depth is reached. Various codes mandate the use cf a separate overflow oiping network in addtion, tothe regula: stormwater drainage system. This network could ecnnect either to the regular system independently outside a building or directly ‘0 the poiat of disposal. Ancther type cf overflow protection is the use of separate overflow pipes or drains, in addition to the regulcr drains, connecting ‘o the regular stormvater drainage piping. Applicable codes and local authorities will provide he information required for design.PLUMBING PPING SYSTEMS c.703 Sidewal! Area Caladlation. . Precipitation falling o1 vertical walls located on roofs ‘and other areas, such as penthouse wals and stax towers, will add to the total runoff calculated fer the horizontal area. Therefore, it is necessary for theie vertical areas to be added to the horizontal tributary ares. To calctlate the amount of sidewall area to adi to the horizontal rocf segment, determine the square foot area of the single or two largest adjacent wal that woud contribute runoff any one roof drain. Divide ‘his area in half, becaase less rain falls on vertical surcaces than on horizontal surfaces. Add the calculated vertical sidewall area to the horizontal tributary area for each roof drain to obain the tol tributary area. Roof Diainage Systems. There are twotypes of systems usedto removestormwa- ter runoff: conventional and limited distharge. The conventional system removes runoff 25 quickly a it accumulates. The limited-discharge sytem removes only @ portion of the runoff, storing the remainder temporarily’ on” the roof. Selection of the appropriate method depends on, the’,capacity and/or availability of disposal faites. and acceptance of the proposed ‘method: by local code and authorites. 95s + wee Roof Diainage Design Procedures" Conventional Roof Drainage Procedur. Designof the conventional mot drain- age syst2m consists of the following general steps: Ae Locase drains on roofs and ancillary areas-threughout allareas of the project discharging into the systems::-< 2. Deiexmine the code overflow requirements. 3, Routt the stormwater piping and overflow systems. 4, Select the rainfill intensity:°This figue:is generilly provided in the code.’ 5. Size the piping network by first calculating th: total tritutary arce for each iridividual drain located in each speciic section’of the rot aid other areas, For horizontal pipe, the sizing procedure starts at the remotes part of the system. Detecmine the pipe pitch aad total tributary arcafor each horizontal pipe section from design pont to design point. 6. Use the appropriate charts to size theroof drain: and the piping network. Enter the charts provided in the applicable code with the intersity, pitch, and total tribu.ary area, Select the figure at the intersection ofthe criteria used for that specific chart, and choose a pipe size corresponding to a figu:e equal toor greater than the calculated tributary arca. Notes associated with different chart types allow conversica to another rainfall rate if different from that of the chart. For vertical piping, use the tributary area disctarging into the vertical leader, increasing the Sze as arezs are added. Table C13.11 is a typical table used to size doth roof drains and the vertical leaders. Table C1312 enables sizing of horizontal pipe: depending upon pitch and tributary area, Table C113 allows the szing of gutters. These tables me from the National Standard Plumbing Code which uses 4 in/h (100 mm/h) 2s the basis for the tatles. Local rates shall always take precedence over a regional value. 7. Wher a very large building, such as awarehouse or factory, has a total tributary area in excess of that appearing in ‘he code, cther methods must be used toe704 TIPING SYSTEMS TABLE C1311 Size of Vertical Conductors and Leaders Maxinum projeced roof erea Diameter ofleader or canductor.t in ff ‘erm 2 Sea 3 2 997 a 3 1,010 0 4 340° 4 5 6200 21 6 10,200 44 8 22,000 93 * Based upon a maximun rate of rainfall of 4 in/h (00 mm/h) and on tbe hydraulic apacities of vertial creular pipes fowing between ohe-hird ‘and one-half fil at termina velocity, computed by the method of NBS Monogieph 31. Where maxinum rates are more of les than 4 in/h (100 ‘un, the figures for draimige area sballbe adjusted ty multplyingby 4 ‘end dividing by de local rain ches per hour 1 The area of rectangulr leaders shall be equivabnt to that of the ‘rear leader ar conductor equired. The rato ofwictd 0 depth of retan- lar leaders sall not exceed 3 to calculate tae volushe of runoff and size of pipe. The mest commonly used method for the calculation of runoff is the rational formula *. Q=AIR . (Ci 1) where Q = quantity (flow rate) of stormwater runoff, ft/s (cfs) (1 ft/s = 448 gpm) A= tributary area, acres (1 aore ~ 43,560 ft” I = imperviousness factor: use a value of 1.0 R = rate (intensity) of -ainfall, in/h (select the rainfall rate based on information contained in the previous discussion) “The design procedure is the same as that destribed for a convertional dranage system, except thet Eq, (C13.1) is used to calcilate the flow rate in horizontal tem piping. Determine the slope of the pipe, and use Table C13.12 to find the pige size up :o the largest arcas in the chart When a larger area is reached, use ‘Table C13.14, entering the chart vith the gpm value and the pitch. Table Ci3.15 Provides a direct conversion in both gallons per hour (gph) and gpm per scuare foot of tributary area for various reinfall rates. For ratesnot shown, add any lesser rate together to find the new rate. In some creas of the country, the public sewer is a combined sanitary and stormwater system. This requires the sanitary and stornwater eystsins to be 2om- bited either inside or outside the building wall. The combined system is usually sized on the 2asis of DFUs. To size the combined system, the nunber of scuare fect of tributary area ofthe stormvater system is converted to DF Js, to caleslate the size of th combined building drainage network. Taole C13.16 can be used as a guide for the conversion of squace feet to DFUs. <PLUMBING PPING SYSTEMS 705 TABLE C19.12 Size of Horizontal Storm D-ains* Maximum projected area fo1 drains of vicious slope: Yin slope Yein slepe Dicmeter of oe cain, in # gem fe pm fe om 3 822 4 1,160 48 1684 6 4 1880 8 24650 no 3760 156 3 3340 139 4720 196 580 278 6 5350 22 7/550 314 40700 445 8 11500 478 16,300 on 23000 956 10 20700 860 292001214 414001721 2 + 33300: 1384 - — ¥7,000 1,953 66500 2768 15 "59500: 2473: M000 3491119900 4946 * Based upon a maxmum rate of rainfall of 4ix/h (100 mami). Where meximurs rates re more or lesethan 4in/h (100 mm), the figures for drainage area shal be adjusted by multiplying by 4and dividing ‘by tho lol rate i inch per howe. LAE = 0.093 mt " Tgpm= 0075.16 Limited Discharge Roof Drainage Systems -., ‘There are tworeascns to select a limited-dischargé roof draiitag> system; voluntarily for economic reasons or by necessity. Tae voluntary use of tae limited-tischarge system vill allow smaller piping throughcut the entire network, resulting in savings ‘on the total installed cost of the Tob conditons tiay mandate the use of a in DN fe) om? gpm Us 3 a0 im” 16 7 05 4 100 30 Ba 5 125 e586 19 6 150 de L480 7 15 13 8 200 «SBD 10. 250 3633515013 “Based upon 4 m 1m rite of rainfallof 4 in/h (100 maf '). Where maximum rates are more or less Una 4 i/h (100 mam 1), te figures for deainage ‘be adiusted by multiplyng by 4 and dividing by the local rate finches per lour ++Gutters other than semicircular may beused providel they hhavean equivalent erosssectionl area. +:Capacities given forslope Wisin shall sed when desning {or peater slopes.c.706 PIPING SYSTEMS TABLE C1314 Capacity of Vertical and Various Sloped Horizontal Sormwater Conductorsand Rainwater Leaders, gpm (W/s) NPS 3 4 5 6 8 19 12 15 DN 30 0 Rs 180 20 250 30 35 Roof drains an¢ vertical piping 22.(58) 12(12) 350 23) 553 (5) 1298 (76) # (1.0) 34 @1) 78 (49) 138 (8) 223 (a4) 479 G0) 88 (54) 138867) 2479 (156) ¥.@1) 8 (30) 1D 639) 19 (12) 35 (20) 67 (43) 120 7) 1983 (123) 350) (220) Horizontal piping slope. ivf (em/m) 4 (42) 8 49 57 (35) 278 (18) 446. (28. 958 (60, 725 (109, 2775 (75. 4958 (312! limited discharge system Urban areas may have overloaded sewers, allowing only a snall portien of the runoff to be discharged into the public sever in order t0 prevent additional overloading, To limit the discharge, install roof drains with factory preset grat» openingsthat allow only a predeterminzd amountof stormwater to enter the drain The allowable amount to be discharged must be given to the manufacturer in order to have the : een * 18) ¢+ TABLE €13.15 Rainfall Conversion Data Rainfall, in/h gph per 1? gpm per1 ft 30 1.870 o.osi 29 1.308 0.02 28 1745 oo 27 1.683 0.028. 26 1621 orn. 25 1558 0.020 - 24 1.496 0.029 23 1434 0.028 22 aga 0.023 24 1.309 0.0218 20 1247 0.0208 19 1184 0.018 18 4122 oo 17 1.060 oor 16 0.997 0.0165 1s ‘0.935 0.0155 14 0873 0145 13 0810 0.0135 12 048. 0.01% ut 0.686 oon 10 0.623 0.0108 gph = 0002s 1 fF = 0093 m?PLUMBING PIPING SYSTEMS e707 ‘TABLE 613.16 Fixture Unit Dranage ‘Square-Footage Equivalent Drainage arca,f? Fixture urit equivalent 180 6 260 10 400 20 490 30 1,000 103 2c00 m 3,000 437 4.000 04 5,000 ™ 7,500 1188 110,000 21500 15,000 2500 20,000. ss yy 3500 . . ©, 28,000, © $300. : Each additional 3 ft? fixture unit : Lf = 0093 m' 1 gpm = 19 vo drain openings sei correctly at the facory. The runoff not dischargec, must be temporarily stored on the roof or on site. . "A siinplified design procedure for te consists of the following general steps! “1 Determine whether the code and locil authors avill permit limited discharge roof drainage: for.the project." 1, vupdie id 2, Establish the’ naximum: allowable flw'rete, in gallons per minute, Sermitted to dischaige into the ultimate point of disposd from all sources of water for the entire site. [f the system is voluntary, some ‘erations will-be necessary after the siping is run to find the optimum pip size for the anount of stormwater stored on the roof. If discharge is intoa sewer, the authorities having jerisdicion will provide the‘allowable discharge amount. If discharge is into a waterway, the existing sit: conditions should te calculated. The difierence between the existing and proposed discharge volumes is tht volume of water that should bbe sored on the roof. If drains in areas other than the roof are present, it is not generally possible to store water from those areas. The discharge Irom these drains wil have to be separated from the total roof discharge t» find the actual discharge allowed irom the roof alone, 3, Decde on'the length of time that vater is to remain on the roof, Generally accepted practice is to allow between 12 and 18 h, staring with is short a time as possible. Lengthen the time as necessary to obtaia the calculated rate of dscharge. 4, Sincz the structural engineer must design the roof to support the additional loadof stored water, a mutual egreenent between the plumbing and structural engiaeers mustbe made to determine the depth If the roof is flat, the generallyc7e PIRING SYSTEMS accepted depth is 3 in. If the roof is pitched, add 3-in,water depth above the high point of the roo’. Table C13.17 gives the weight for each inch of watcr, TABLE C13.17 Weight of Rainfall Amcunt of rain,in Weight of water, Ib? 6 31.21 5 26.01 4 20.81 3 15.61 2 10.40 oot 5.20 in = 254 me IbIf? = 4.88 kg/m? 5, Find the total amoust of rainfal, in inches, that will fall on th: roof for the * cime established in step 3. The cmount is obtained from Fig. C13.24, whic is 2.24-h, 10-yr return reinfall.’A 10 percent reduction ir the 24-h fgure approxi- ‘nates an J8-h rainfall. 6. Divide the figure found in step 5 by the time found instep 3 (e.g, 8 in divided dy 12 h equals 0.67 in/h). Thea determine the gpm discharge, using Table £13.15. Compare this figure-witn the maximum allowable discharge from the roof foundin step 2. Adjust the retention time as required to matek the allowable discharge rate.“ ' Shae ty whe Consideration must. be given fcr a heavy rainfall occurring afier the’ dedign storm ends, while sone watcr Yemains on the roof during the drain-down tine, This rainfall should te the actual 1-h duration, ‘with the’ same return period selected for the project. This 1-h rainfall will deposi: a number of inches of vain. The actual drainage rate will be the figure selecte¢ in step 2. By subtracting che drainage rate figure (in inctes per hour) from the 1-h rainfall figure, the actual tota. allowable rainfall depth is found. This death must not exceed the depth established in-step 4. If it does, additional depth of storage is required, or the client must be willing to eccept rainwater spilling out of tre emergency overflow oace during the return period selected, 8. Locate drains on roof, and find he total number of crains requred. 9. Route the stormwater piping and overflow systems. * Divide the number of drains into the total roof discharge allowed to find -he 3pm flowirg out of each drain, Using Table C13.14, size the individual dreins and branches, using ‘he gpm and pitch of the pipe. Limited-cischarge roof drains are set at the ‘actory for the specified maximum flow raie established by the engneer. Size the pining network based on accumulated gpm fow at each design pcint and the pitch of the gipe. 1. 5 1m. Conventional Roof Drainage System Considerations L Mis importent that the local codes and authorities be consulted for the method sed to comect to the public sewers. °