r : Gaswelll | conn Sul Northern Gas Fipeliaes Testing); “~ Theory, Practice & Regulation David A.T. Donohue President. Mic, ‘Arlington Exploration Company Turgay Ertekin Assistant Professor. « The Penesytvania Sate Unverty « Daperimental-Library ” age PetrstUMm & Gas Eng” Wh the assistonce of Roward 8. Wriversivy of tage: & Technology’! mardB. Bradley AN ORE. gy ie . Be IntornationdRuhda esemg Copyright © 1982 by Intemational Hina Tt Development ortimtion. Alt ighis reserved. No past of Us ‘book may be used OF Serge in any manner whateaever withow waitten permission of Tee publisher except in the case of brief quote ‘embodied in «itical tae ed revews. For information address: RDC. Publishers, 137 Newbury Set, Boston, MA 02116 198N: 0-934634-10-6 Cloth ISBN: 0-934634-12~2 Papet Lira of Congress Catalog Card Number: 8180726 Printed in the United States of Americajor of al 137 nag: University of fisy G Teehnctogh LAH OME-31 Table of ContentsAcknowledgments Preface Introduction (Drconventional & tsochranal Tests 1.4 Derivation of Fundamental Flow Equation 122 Conventionat Test 13 bbochronal Test 1a Maiied sochrons! Test (Bitton teins Onin 2.1 Pscudopressure (Real Gas Potential) Treaiment for Gas Flow 2.2 The Radiat Diffusivity Equation for Gas Flow 2.3 Analytical Solution of the Diffusivity Equation 2 ow Regines 25 Shin aco 2.6 Turbulent Flow Factor 2.7 Drawdown or Falloff Test 2.8 Buildup Test 2.9 Determination: of 5, Sand Dfrom Larldup ant Gendt 3. Additional Testing Considerations ‘3.1 Sutface-Pressure Relationships 32) Stabilization Time andl Radius of investignion 33. Choice of Flow Rates and Duration of Flow 3.4 Wellbore Storage Effects 3,5. Production of Liquids 316 Single-Point Test 3,7 The Orifice Meter 3.8 Testing Equipment 315. Measurement, Sampling, and Safety a Te 4, Procedures & Regulations i pteatate Onl Compact Commission Proce’ 4.2 Stale of Texas Regulations 43 State of Oklahoma Regulations 1. State of New Mexico Regulations 45 State of Louisiana Regulations “£6 Province of Alberta Regolations vi Gaswell Testing 24 3 B uM 6 40, at n a” 53 34 02 64 65 66 68 68 74 76 81 83 102 109 123 1245, Field Test & Interpretation 5, Exercises Nomenclature & Ute ‘Appendix. A Derivation of the Conlinuily Equation Appendix B Analytical Solution of the Dilfusivity Equation Appendix © Skin "0 to Restrict} Entry ‘Appendix 2 cov Deterr nation of Gas Supercompressibility Factor and Viscos:'y Appendix Sample Data porting Forms Index 127 133 147 151 135 “165 WwW 185 199 209 ‘Table of Contents viiAcknowledgments vill Gaswell Testing The authors of this book wish to thank the following individuals {or the active support and encouragement they jnided during. is several stages of completion: Howard 8, Bradley of Mobil Exploration and Producing Services, Inc. in Dallas who reviewed the full manuscript including the tables, figures and equations, offered continued support and cti- {icism, and whose personal interest in both video production and the book led to its satisfactory completion; Aziz S. Odeh and William Prachner of Mobi Oil Corporation, Dallas, for the valuable suggestions they offered during the plan. “hing and working stages of the book and video production; Bruce Jankusa of Mobil Oil Corporation, Lake Charles, Louisiana for advice during planning and for his willingness to appear in the video production; Grant johnson of Mabit Canada ttd., Edmonton and R.G. fustice of Mobil Oil Corporation, Denver for their advice and secoin- mendations during writing: Derek F. Harvey of Mobil Oil Corporation, New York for his care- ful helmsmanship; and especially to Michael Hays and Gail Smith Of THRDC for editorial assistance, review, and final publication.lviduals uring its This hook accompanies a videotape program of the samo name. ‘The combined \:‘:otape and book, referred to as a module of instruc- tion, was one of three prepared by IHRDC on a join Inasis with Me Oil Corporation during 1980. The three modules, one each in geol- ay. gecnhysics, and petroleum engineering, were produced to deler- imine whether this medium of instruction would provide an effective ‘way of teaching recent graduates and those individuals changing spe- aration, alts, “what they nced to know, when they need to know it.” A. re plan reat deal was learned during the pilot stages, Properly designed and vs Properly used, video-assisted instruction was found (0 be elective, uisiana elicient, and convenient. rin the eit the: provides one to train young graduates Justice IFRDC sought financial and advisory lecom support from a limited number of companies to undertake the develop. ‘ment of tse Basic Technical Video Library for the E&P Specialist. To scare date the following companies have agreed to serve as Sponsars: Mo- t Savithy bil, AGIP, ARAMCO, Cities Services, Dome Petroleum Utd., Gulf, tion. Philip, Stanel:t Oil of Calfornia/Chevron, and Texaco. Work on the Library began in July 1981, With an accelerated production schedule of 24 modules per year and the continued sup- port of the Sponsors, the Library shauld be completed in about five years. Preface ix _Introduction| Figure Lt Schematic of Gast Sytem, 2 Gaswell Testing Gas wells are tested to determine their ability to produce gas under various conditions of surface or sandface (ndface pressure as used here is the same as thr hottomhole tubing pressure) and reservoir pressures. The controlling surface pressure will depend on the operating pressure of the pipe- line to which the gas is delivered, the characteristics of any installed compression, the pressure losses that occur in gathering lines, and the choke size selected to control the production rate. Also, pressure losses that occur in the tubing must be added. The reservoir pressure depends dn the extent of reservoir depletion and on any pressure sup- port that might be provided by water encroachment. A typical flow system is shown in figure 1.1. Several general conclusions may be drawn from figure Wa: revere Lomes in ‘Presse Gain Through Compression [+ pveaire Lots in Tubing |-a a Bp (cee pres) Flow! = Flow Low (sande pressure]ndface (1) Maximum delivery will occur when the reservoir tubing pressure, Bp, is a maximum and the sandface pres- urface . sure, Pf, is. a minimum. pipe- 2) A high pipeline pressure, without installed com- of any pression, will reduce deliverability. cur in (3) Pressure losses in the gathering system and well- ol the bore will reduce deliverability. in the (4) Installation of compression will increase deliver- nds dn ability. e sup- (5) For a given sandface pressure, deliverability will ent. A decrease as the reservoir pressure is reduced through depletion: figure (6) The fotmation characteristics of wells are different, = therefore, each must be tested to determine its unique ability to produce. Because the capacity of a well to produce into a well- bore isa unique characteristic of the well (its subsurface and surface flow system), field testing procedures have been developed that allow the engineer to predict the manner in which a well will produce over its reserve's life under the various operating conditions that may be imposed. The results of these tests are used by the engineer to: © Determine whether a well will be commercial. ® Satisfy the regulations of state agencies. 4 Determine allowable gas production rates. Design processing plants and pipeline extensions. ® Serve as a basis for gas sales contracts. © Support cs liverability and compression studies. ® Determine well spi ing and field development pro- gras ® Plan cycling programs. ® Determine the need for stimulation, ™ Assist in the identification of reservoir boundaries such as faults, pinchouts, and water tables. Introduction 3a 4° Gaswell Testing Accordingly, it is important that the engineer understind fully the background, procedures, and interpretation of gaswell tests. ‘The purpose of this book is to cover the principal as- pects of gaswell testing. The chapters follow a topical sequence: chapters 1 and 2 are devoted to the theoretical fundamentals of testing; chapter 3 to test design considera- tions; chapter 4 to test procedures and state/province regulations; chapter 5 to the procedures followed and interpretation made for a specific gaswell test in Louisiaina, and chapter 6 to exercises, Mathematical developments, correlations, sample reporting forms, and similar material are given as appendixes.Chapter} |1 & isochronal TestsDeliverability tests on gas wells have been used for many years to determine flow capacity. In early times gas wells were rated by opening the well fully to the atmosphere snd i] 2 . : determining its open flow capacity. in order to prevent wastage of gas and possible formation damage, open flow. potential (capacity) tests of gas wells were soon determined. by flowing the wells against a particular pipeline backpres- sure, Since a backpressure is maintained at the surface} during these tests, they are commonly called bach ivssure tests. Lt Derivation of Fundamental Flow Equation In order to understand the relationships that exist when g flows through a formation into a wellbore, certain mathe matical developments are necessary. We start with thd derivation of Darcy's law for the radial flow of gases. Co sider radial flow toward a vertical wellbore of radius 4, in horizontal reservair of uniform thickness and permeability as shown in figure 1.1, The radius at the external boundar is rand the pressure is 2. If the flow rate of the gas at any radius, r, sq it can be shown that vn(Se)oa (re) te te where Gy = 2nshk at r= fy, ie., the flow rate at the Mz By Darcy's Law: 6 Gaswell Testingmany wells re ond revent n flow mined kpres- Pe uiface ossure Figure 1.1 Schematic snd Nomercatie oe 2 Radial ow Sytem dr= Pri Py 2) be dr al standard conditions wtie en gas. . . rahe, a tebe, = Gc GP G30 th the a? : Con- and iyina ability te PB indary ac f (4-5 ) = 2ahk | pdp (1.4) ao roe Zz : can be fw Ms Pod Integrating gives: ay acCUin B= 12) = auto. 2 — pee) or we, ity the . 0,703 x 10°*khip? — Per) 1.5) é ey k\oiny / TOE) Conventional & Isochtonal Tests 78 Gaswell Testing The equation (1.4) can be integrated from py to p to obtain the pressure distribution as a function of r. The average reservoir pressure, fe iy Pat For (6) Substituting for y#.from the previous step, integrating and tearranging gives: 0.703 -x_10-*khiy de = But) a, Te (” fu In equations (1.5) and (1.7), the pressure-depcndent param-| elers of viscosity py and compressibility factor z are assumed to be evaluated at the average reservoir pressure jy and therefore treated as constants. Now let CaO 2OR IONE kh al Bele 2 («" (5) - ors Then dsc = CUB ~ Pu?) (19) where average reservoir pressure, psia flowing bottomhole pressure at fw, psia permeability of the porous medium, md. formation thickness, ft reservoir temperature, °R gas viscosity at average reservoir pressure, cp!10 pio ir. The 25 compressibility factor at average reservoir pressure, fraction external radius, ft well radius, ft production rate, MMSCF/D performance constant for the well wa (6) fy Ge ¢ ling and Since the ratio fe/iw in equation (1,7) and others to follow is usually very large, the — 0.75 value is sometimes omitted in Looe a7) deriving the equation. Taking the logarithm of both sides of aoe equation (1.9) param- tog Ge = log € + loge ~ pu) (1.10) rz are ssure Pre This equation indicates that for Darcy-type laminar flow, the log-log plot of q- versus (fx? — Py?) is a straight line whose slope is 45°. Because of deviations from ideal flow behavior, the actual plots may be nonlinear and their slopes fess than 1.00. Rawlins and Schellhardt,! drawing upon ‘empirical observations, modified equation (1,9) by adding the exponent n to account for the observed deviation from the ideal flow behavior, The deviation is due to several factors, the most important being the effect of turbulence. Thus, equation (1.9) takes the following form. “(1.8 Ge = Cle — Pat'l® aay The range of values for the exponent nin the above equation varies between 1.0 and 0.5, approaching 1.0 and 0.5 at complete laminar and turbulent flow conditions, respec- tively. The logarithm of both sides of equation (1.11) yields, log qc = log C + logit ~ Pye) (1.12) From the definition of performance factor, , it is obvious Conventional & isochronal Stes1.2) Conventional Test 10 Gaswell Testing that this factor depends on gas properties, such as viscosity, compressibility factor, and reservoir properties, such «i permeability, temperature, formation thickness, external boundary radius, and wellbore radius. Looking at equation (1.12) it is seen that a log-log plot o (Be ~ Pa?) versus dic should give a straight line wh slope is 1/n. Inorder to obtain a straight line we need at feas {wo points through which the line may be drawn. In general practice it is suggested that four points be obtained to min: mize error. @# [fo perform a conventional test, the static reservoit pressure, py, is determined by shutting in the well"foF period of time, until its stabilized pressure is known, A flow fale, Gq is then selected, and the well is flowed until ig lowing pressure stabilizes. In many areas, stabillzatfon ‘defined in terms of percentage change per unit of time.-Thig Mabilized flowing pressure, yi, is recorded: This yr0R dure is repeated four times by sequentially decreasing oj preferably, by increasing the flow rate and measuring th corresponding stabilized flowing pressure each time. flow rate and pressure histories for a conventional fest a shown schematically in figure 1.2. Now, we can plot the point (fj? — Pa) Versus de each measured rate and draw a best-fit line through 1 points as shown in figure 1.3. I The straight line of figure 1.3 has a slope 1/n and known as the backpressure line or inflow performance ling Trom this straight line and equation (1.12) one can dete mine the absolute open flow potential or AOF, the theoref ical flow rate that would occur if the sandface pressure wal,scosity, uch sis, external Figure 1.2 Fle Rate and Presse History ofa Tpical Comeninal Tes. gh th slope = Ve nand i: nce line] in deter. theoret wa! 1 16 Oo sure Figure 1.3, aye (MMSCFIO) —> For Conca WH re Eamoe, ie Conventional & isochrone AL| Example 1. Deliverability Calculation for a Conventional Test 12. Gaswell Testing reduced to zero, Having obtained values of a and C, eq tion (1.11) may then be used to calculate the deliverabil {inflow rate) of a well for any given set of sandface aq reservoir pressures. ‘The following example illustrates the application of equation (1.1 Suppose that we know that C= -0.0037 (for qin MMSCFID): n= 0.93, what is the flow rate when Be = 3000 psia Pat = 1850 psia? ae = CHE = Bl Ge = 0.0037 (3000 = (B50¥r" Ge = 6.96 MMSCED What is the AOF? AOF = 08037 ((3000)' = (our AOF = 10.86 MMSCF/D Itshould be recognized that only a single flow-rate! is needed in order to obtain the backpressure flow equatl if the value for nis assumed or known. This is the case, ff example, for the single-point test allowed by the Oklahog Corporation Commission—~in that particular situation, § commission allows the operator to assume a value for 0.85. Although the conventional test requires four. cons flow rates, itis highly unlikely in practice, that a flow will remain constant. Because of wellbore storage efi and because sandface pressure changes occur, the flow: during each flow period will decrease with time as show figure 1.4. Winestock and Colpitts,” to account for variation in rate, suggested that the instantaneous flow (say at the end of a flow period) be used rather thanC, equ iverabilit dface ar ion (1.12), ASCH), 0 psia ai “average flow rate. In order to incorporate variable rate flow tests, they suggest the use of normalized equations. Conventional gaswell testing and the interpretation of its data are relatively simple. Although the test has been con- sidered the basic standard for many years, it has: certain drawbacks. (The difficulty arises when the reservoir per ‘meability is low, or flaring is to be minimized. In this type of reservoir a properly stabitized, conventional deliverability test may not be conducted in a reasonable period of time. In other words, the time required to obtain stabilized flow conditions may be very long.) Callender? proposed the isochronal gaswell te inwhich a well is shut in long enough before ea: ‘Petiod so that each flow will begin with the same pi distribuifon in the reservoir. The basic principle behind isochronal_tecting is that the effective drainage radius,_a {aifefon of the duration of flow, is the same for.each mea- st. atest Figure 1.4 Schematic Showing Actual Flow Rate during Tex. 1.3 ’ Isochronal Test = * Conventional & Isochronal Tests #13a Figur Test Fle ined Presse: Hise of an Batu 14° Gaswell Testing sured data point, Thus, separate flow tests, run for the sar length of time, will affect the “same radius of drainag Therefore,_each_performance.curve_(r_line_on_a Tog-ld for a fixed radius of drainage. Cons bbut the yalue of the performance Corista , will depend on the duration of flow (ocraditis ofa age). An accurate value_for. C.will be obtained.with a flo test that reflects the behavior of a large radius of drai often a long-time The procedure normally followed in this testis outli below (also see figure 1.5). (15 The well is initially shut in until the static pressuq _-7 Bry Slabilizes. (2) The well is opened to production at the first ral Ge, flowed for a predetermined and fixed period + time, and the bottomhole pressure is measured; (3) Steps 1 and 2 ate repeated two or three additio f oe Ge (mnasce/Ottimes at different flow rates. The duration of each flow period should be equal and the pressure after cath flow period must build up to its original static value. The duration of buildup, then, will not nar- A mally be equal to that for the flowing periods. (4) Finally, after the jast flow period, one flow test is conducted for a time period long enough to attain s‘abilized flow conditions. This period is usually called an extended or stabilized flow period. The rate of flow during this last flow test need not be equat to that of the fourth isochronal rate, and, in iact, i$ often less; The time required to reach stabil- ized ‘pressure wilt depend on the properties of the first rai formation and the fluids. : Iperiod ; , “a sasured. ad 1¢ isochronal flow data that are obtained following ~editiong the above outlined procedure are analyzed as follows: MF). \ti) The three or four isochronal points* are plotted on . lop-log paner as shown in figure 1.6. UD the best swaight line is drawn through the ‘sochronal points. LASS The value of the exponent n is obtained from the slope of this line. Remember the inverse of the slope will give the value of n. The value of a, “4, btained from the isochronal test, should be equal to that obtained by the conventional test; however, as explained earlier, the position of the line and, thus the value of Cis not correct. (A) The stabilized deliverability line line withthe cor- rect value of C) is obtained merely by plotting the stabilized point representing the extended flow rate and stabilized flowing sandface pressure, and drawinga line through it parallel othe plotted line. x Bap aoe ca be we tt efit pais iwo Be pe ~ Path oxi" ‘st Stobiizes : Ontverabiity 19 E fntow! Line Tranaient Delerabitty opes Unf bine Intercept 7 Oe Ue - Figure 1.6 10 Leyes Plt Typical hoelvanal Wel Test Aq HSCF) — 15) Provided there are no significant turbulent flow pressure losses at high rates (see section 2.6 for correction for turbulence), the AOF or the flow rate abtained against any sandface backpressure can be easily read from the stabilized inflow performance line. Example 1.2 Deliverability Caleulations for \ an Isochronal Test Analyze the following, isochronal well test data. « ~Dovaion a Few ~ ~~ Sandindé —— iow Tate Test thors) Pressure (pda)_OMMISCEY) Shula 7200 % . © ies 2a ® 782 as 6 16a? 45 6 1511 54 16 Gaswell Testing‘Alierwards, the well continued to produce at 6 MMSCF/D and reached a stabilized flowing sandface pressure of 1180 psia Solution ‘The following table is prepared: Bellis) ae NSC) pay — a nes 8 1992 2a 358 x 10" 126 x toe va 34 3.18 108 155% 108 1647 ‘40 27h x 1 21) si 54 228 x 10° 256 x 1 Sabir Point * rT 6.0 139 «108 345_x 10" Plot (Bi? — py?) versus qu: on log-log paper. The slope of the deliver- aby Ln og iS 10) = tog 0.43 4109 op 2 7 Togo" % (0) = logti > Toy : = 094 C= —% » 6 __ a + = Be Gas x 10H C= 429 x 10+ ‘The AOF from the plot is 8.0 MMSCF/D. See example 1.2 figure. By’ calculation its AF = 4.29 x 10 (2200 — OI" = 8.25 MMSCHD e Coniared to the conventional test, the isochronal test should save a considerable volume of gas from being flared 14 . Modified.” Isochronal-Test sees1. AOF = 8 MMSCF/O 1 10 100 Example 1.2 Por of Baa for Example 1.2. Age (MMSCF/D) —> into the atmosphere. it may also save time if the buildup to slatic pressure after each flow period is relatively short. This . time saving during the flow periods may be considerable in the testing of wells producing from tight gas sands(in ex- tremely tight gas reservoirs, an isochvonal test may not always be practical since it is very difficult to attain a completely stabilized static reservoir pressure before the initial flow period and during each subsequent shut-in period. + |? In 1959, Katz et al. suggested a modification to the isuchronat test. They suggested that both the shut-in period and the flow period for each test could be of equal duration provided that the unstabilized shut-in pressure, Pye, at the end of each test be used instead of the static reservoir 18 Gaswell Testingpressure, pp, in calculating the difference of pressure squared for the next flow rate. Figure 1.7 outlines the flow tate and pressure sequence of a typical modified isochronal test. Extended Flow Rate gq (MMSCE/O} 2 (oa) __, The analysis of modified isochronal test data is similar to that shown earlier for the isochronal test. As in the isochronal test, the first four points are plotted on log-log. paper as shown in figure 1.8, It is important to remember that the associated unstabilized shut-in pressure is used tead of stabilized static pressure in calculating the differ- ence of pressure-squared for Ihe next flow rate, A best-fit line is drawn through the four points to determine the slope and a second line parallel (o the first is drawn through the stabilized point, The values of n, C, and AOF are found in the same manner as outlined in the conventional and isochronal gaswell testing procedures. Figure 17 Flow Rate and Pesture History fora Typkeat ‘Meret ochronal Tes. 4 Conventional & Isochronal Tests 19.Delivrabiity i : Stapiized intl rate) Line “ranent Deters Tae tine ear m 1 , \ i Figure 1.8. oy io 100 Lig Pot Byes Meet obo j Gane Hs ye (MISC 0}— | Exampte 1.3 Deliverability Calculations for a ified Isochronal Test Modified Isod Analyze the following modified isochronal well test data, ( in of Sant Test hour)___Pressure (sia) (ASCO) i 4 7000 y wa i 10 3982 10 wie 0 1360 10 isp 10 1913 10 1306 2% 1072 : £8. 202. Prepare the following table. 2 Gaswell Testing” ,me BCT i — 4 bal x 10 —— 69 "oa x a ° —~ ne Lin aoe 6 . roo Las x r08 —~ 190 2.03 x 10! ei = fe o. Stabe point Plltiwe’ = Pact -- + yan log-fog paper. We can read the slope ofthe deliveratli., oan: os Fallows: LL log (2.85 x 109 = log (o.14 x 109 hat 6 ° Tog 10 x 109 = Tog(l x 10) 100 orn = 0.%6 0 data. a’. . mt! #10 * ‘pau? 0 ADR» 19 MMSCE/D 1 0 19 Beample 1.3 gq MMSCF/O) —e- Plotod Dato Ecmpe 13. Conventional & isochronal Tests 21References 22 Gaswell Testing ‘Then constant C, can be cafculated as: 4, =~, 10 Pa) (2.85 * 104" Gee = ¢ © = 0,000124 From the plot shown in igure 1.8 wecan read thy SOF asbeinged 13 MMSCHD or alternatively, AOF = Clpwat ~ (0)"1" Aor = 0.000124 (2000)! ~ (01 AQF = 12.91 MMSCHD which is quite lose to our reading from the plot "Rawlins, EL. and Schelihardt, M.A.: Backpre.sure O Natural Gas Wells and Their Application to Production P Monograph 7, U.S. Bureau of Mines, 1936. 2 Winestock, A.G. and Colpitts, G.P.: “Advances in Est Gas Well Deliverability,” J. Can. Pet. Tech. duly-Septembergy ME-9. 3 Culfender, M.H.: “The Isochronal Performance Method termining the Flow Characteristics of Gas Wells," Trans., Al 204, 137-142. 4 Katz, D.L., Comell, D., Kobayashi, R., Poettmann, FH J.A., Elenbaas, .R., and Weinaug, C.F,: Handbook of Nat Engineering, McGraw-Hill Book Co., Inc., New York (1959),