Professional Documents
Culture Documents
com
DIGITAL TRANSFORMATION
Successes and challenges
installing the first proprietary
closed-coke slurry system
24
EDITORIAL
Executive Editor Adrienne Blume
Managing Editor Mike Rhodes
Digital Editor Stephanie Bartels
Technical Editor Sumedha Sharma
Reliability/Equipment Editor Heinz P. Bloch
Contributing Editor Alissa Leeton
Contributing Editor ARC Advisory Group
Contributing Editor Anthony Sofronas
4 FEBRUARY 2021 | HydrocarbonProcessing.com
We understand how you need
to reduce complexities at your plant.
CLEAN PROCESS
+ CLEAR PROGRESS You strengthen your plant’s safety, productivity
and availability with innovations and resources.
• For raw and processed natural gas, coal seam gas, shale gas,
gas mixtures, wet biogas and digester gas
• Powerful process control thanks to pressure and temperature-
compensated values measured in real time
• Advanced gas analysis functions for calculation of additional
process variables, customer-specific gases or gas mixtures
• Maximum reliability and robustness independent of the
gas moisture
16 Executive Viewpoint.
from the adoption of AR/VR. These ded with AI, have also been used to
technologies enable plant personnel to monitor safety hazards (e.g., corrosion) Hydrocarbon Processing spoke
better train new and seasoned engineers in processing plants and drill rigs. These with Andrew McCloskey, Chief Technology
for various field work, emergency situa- AI-embedded inspection bots can learn Officer and Head of R&D at AVEVA,
tions, plant disruptions and/or mainte- to “talk” to each other and communicate about how digital twins and cloud-based
nance functions. Wearable AR headsets issues regarding potential safety hazards services are used, their benefits and
enable the user to view specific data on that need maintenance work. challenges, security issues, etc., as related
various pieces of equipment. This is done to the downstream oil and gas, refining
by projecting digital images on the head- Cybersecurity. The use of new digital and petrochemicals industries.
gear’s lenses. Maintenance personnel can technologies has also opened the HPI to
simply look at any piece of equipment,
and the headgear lens will display metrics
such as flowrate, pressure, etc.
additional cyber threats. This predica-
ment has led organizations to invest in ad-
ditional cybersecurity technologies to pro-
24 Digital transformation.
The adoption of new digital
technologies is changing the way the
tect critical infrastructure and company/ HPI operates. These enhancements are
Artificial intelligence (AI). AI is being employee data. These cyber-barriers (e.g., enabling producers to operate more
used in software and digital platforms to firewalls, encrypted data) are crucial to efficiently, safely and more profitably.
monitor equipment health and provide keep plants, companies and workers safe. This month’s Special Focus section details
users with predictive analytics, which several areas where digital transformation
is decreasing equipment downtime, in- Plant design, engineering and con- is having significant impacts.
creasing reliability, providing actionable struction. Finally, new digital technolo-
diagnostics for enhanced plant/unit per-
formance and much more. AI can also
gies are aiding the capital construction
industry. Digital tools (e.g., software, AR/ 41 Process Optimization.
Grupa LOTOS implemented a
proprietary closed-coke slurry system
boost energy efficiency, monitor plant VR, modeling systems) are enabling EPCs
emissions, optimize production and sup- to better design, engineer and configure (CCSS) at its facility in Gdansk, Poland.
ply chains, and substantially increase mar- individual units, plant infrastructure or This article provides a comparison
between a conventional coke handling
gins and profitability. whole grassroots facilities, as well as ex-
system and the proprietary CCSS
pansions, unit integration, commissioning,
technology installed at the facility, as well
Digital twins. A digital twin is a digital supply chain, logistics and so much more.
as details of the successful integration of
copy of a system, unit or process that can
the CCSS in the delayed coking unit.
replicate real-life operations. Digital twin Takeaway. New technologies are trans-
75 Safety.
technology is helping provide companies forming the way plant owners/operators, Operational discipline
the ability to simulate operations and con- EPCs and service/supplier companies are is a fundamental part of
figurations on units—to conduct “what optimizing operations. This issue’s Spe- effective programs for achieving
if ” scenarios—without affecting real-time cial Focus is devoted to the advancement excellent performance in process
operations. The technology is crucial in of these digital technologies and how they safety. Steps for getting started in
providing operators with a virtual “best are propelling the HPI to increased reli- implementing an operational discipline
path forward” for unit efficiency, as well ability, profitability, safety and sustainable program, or improving an existing
as a beneficial asset for training personnel. operations. effort, are discussed here.
Business Trends
10 FEBRUARY 2021 | HydrocarbonProcessing.com
OH YEAH,
WE HAVE A FILTER FOR THAT.
No doubt about it. Whatever process you’re running, we have a filtration and separation
solution that will save you time and money in the long run. And if your process design
has you stretching out into new territory, don’t worry. New territory is kind of our thing.
713.849.0849 | ftc-houston.com
LEE NICHOLS, EDITOR-IN-CHIEF/
ASSOCIATE PUBLISHER
Lee.Nichols@HydrocarbonProcessing.com
Global Project Data
Hydrocarbon Processing’s Construction Boxscore Database is projects are in preconstruction phases. Projects within the planning/
tracking nearly 1,400 projects around the world. The Asia-Pacific proposed phase represent the largest market share for active projects
region accounts for 37% of active projects market share, followed by (37%). More than 40% of projects in the planning/proposed phases are
the U.S. (17%) and the Middle East (14%). Approximately 67% of active within the Asia-Pacific region.
10 8 23 61 47
41
Canada
Eastern Europe,
25 30 22 Russia, CIS
98
62 70 Western Europe
116
U.S. 56 203 193
25
61
Middle East 117
22 23
Refining 35
Petrochemicals 13 9 Africa
Gas processing/LNG Latin America Asia-Pacific
32
28 28
19% Engineering
25 26 26
33% Under construction
20
18
15 15
17 5% FEED
13
11 11
37% Proposed/planning
6% Study
Nov.- Dec.- Jan.- Feb.- Mar.- April- May- June- July- Aug.- Sept.- Oct.- Nov.- Dec.- Jan.-
19 19 20 20 20 20 20 20 20 20 20 20 20 20 21
Detailed and up-to-date information for active construction projects in the refining,
gas processing and petrochemical industries across the globe | ConstructionBoxscore.com
12 FEBRUARY 2021 | HydrocarbonProcessing.com
June 2-3, 2021 | IRPC-PROCESS.COM
Process
Optimization
and Transition
Technologies
Why you need to attend:
IRPC-PROCESS.COM #IRPCPROCESS
MIKE RHODES, MANAGING EDITOR
Mike.Rhodes@HydrocarbonProcessing.com
Industry Metrics
Global refining margins showed mixed results. Refineries have returned Global refining margins, 2019–2020*
from the autumn peak maintenance season, showing a slight rise in available 25
spare capacity globally. The U.S. saw improvement in transport activity during WTI, U.S. Gulf
20 Brent, Rotterdam
year-end holidays, but refinery intakes remain around 2 MMbpd below y-o-y
Margins, US$/bbl
15 Oman, Singapore
levels. European markets weakened, affected by stronger crude prices amid
seasonal weakness. Asia refining margins suffered mostly in the bottom of 10
the barrel due to a decline in fuel oil requirements from the utilities sector. 5
An expanded version of Industry Metrics can be found 0
online at HydrocarbonProcessing.com. -5
Dec.-19
Jan.-20
Feb.-20
Mar.-20
April-20
May-20
June-20
July-20
Aug.-20
Sept.-20
Oct.-20
Nov.-20
Dec.-20
U.S. gas production (Bft3d) and prices (US$/Mft3)
120 7
Global refining utilization rates, 2019–2020*
100 6
100
80 5
90
Utilization rates, %
4
60 80
3
40 Monthly price (Henry Hub) 70
2
20 12-month price avg. U.S. EU 16
Production
Production 1 60
Selected Asia
0 0 50
AM J J A S OND J F MAM J J A S OND J F MAM J J A S OND
Dec.-19
Jan.-20
Feb.-20
Mar.-20
April-20
May-20
June-20
July-20
Aug.-20
Sept.-20
Oct.-20
Nov.-20
Dec.-20
2018 2019 2020
Production equals U.S. marketed production, wet gas. Source: EIA.
Selected world oil prices, U.S. $/bbl U.S. Gulf cracking spread vs. WTI, 2019–2020*
100 50
90 40 Prem. gasoline Diesel
Cracking spread, U.S. $/bbl
June-20
July-20
Aug.-20
Sept.-20
Oct.-20
Nov.-20
Dec.-20
Dec.-19
Jan.-20
Feb.-20
Mar.-20
April-20
D J F M A M J J A S O N D J F M A M J J A S O N D
2018 2019 2020
World liquid fuel supply and demand, MMbpd Rotterdam cracking spread vs. Brent, 2019–2020*
110 8 30
Implied stock build World supply Forecast
105
Stock change and balance, MMbpd
6
Supply and demand, MMbpd
Jan.-20
Feb.-20
Mar.-20
April-20
May-20
June-20
July-20
Aug.-20
Sept.-20
Oct.-20
Nov.-20
Dec.-20
70 -4
2016-Q1 2017-Q1 2018-Q1 2019-Q1 2020-Q1 2021-Q1 2022-Q1
Source: EIA Short-Term Energy Outlook, January 2021
Singapore cracking spread vs. Dubai, 2019–2020*
Brent dated vs. sour grades 20
(Urals and Dubai) spread, 2019–2021*
Cracking spread, U.S. $/bbl
10
60
0
Light sweet/medium sour
Dubai
crude spread, US$/bbl
40 Urals -10
Mars
20 -20 Prem. gasoline Gasoil
Jet/kero Fuel oil
-30
0
Dec.-19
Jan.-20
Feb.-20
Mar.-20
April-20
May-20
June-20
July-20
Aug.-20
Sept.-20
Oct.-20
Nov.-20
Dec.-20
-20
Nov.-20
Dec.-19
Jan.-20
Feb.-20
Mar.-20
April-20
May-20
June-20
July-20
Aug.-20
Sept.-20
Oct.-20
Dec.-20
Jan.-21
14 FEBRUARY 2021 | HydrocarbonProcessing.com
HEINZ P. BLOCH, RELIABILITY/EQUIPMENT EDITOR
Reliability Heinz.Bloch@HydrocarbonProcessing.com
flexibility and agility, as well as the abil- tomers the type, size, connection source, HP: What does the future
ity to collaborate seamlessly from their operational background and location of refinery/plant look like?
technology deployments, and AVEVA is equipment installed on a plant site. This McCloskey: The future refinery/
responding to this need. data is generated in capital projects, from plant will likely put sustainability at the
new plant builds to brownfield revamps forefront: managing energy usage and
HP: How does the current volatility and retrofits, and forms the backbone of costs, optimizing process yields, and re-
in oil prices impact your customers? the digital twin. ducing or eliminating safety-related inci-
What are the challenges associated Lowering total cost, time and risk dents will contribute to this goal.
with using digital twins? in capital projects are the major chal- The industrial world is progressing
McCloskey: The current macro oper- lenges associated with implementing through the transformation cycle that
ating environment for oil and gas compa- digital twins. While each industrial en- retail and finance experienced 10 yr ago.
nies has accelerated the need to optimize vironment brings different challenges, We are working with leading companies
manufacturing operations and improve the petrochemical industry is well-po- in energy to increase operational effi-
performance to protect profitability. sitioned to adopt digital twins. In fact, ciency, unify their data and connect their
Ensuring the accuracy of data is criti- early adopters of digital twins have pre- teams to realize Industry 4.0. Companies
cal to success. In response, we are in- dominantly been the continuous process like BP are using us to accelerate deci-
creasingly seeing these producers invest and certain discrete repetitive industries, sion making from weeks to hours and
in their own cloud-based data platforms which have higher levels of automation. optimize profitability throughout their
for current and future capital projects, Due to their highly instrumented envi- value chain.
operations and maintenance as part of ronments, these industries can more eas- The pandemic has accelerated our
their digital transformation projects. By ily minimize essential workers onsite and conversations because everyone is un-
and large, this shift is due to the fact that shift many operational functions offsite der pressure to cut OPEX and minimize
accurate data, kept in one place, ensures using high-fidelity digital twins. CAPEX and risk. Software can materi-
the reliability of a digital twin’s output Undoubtedly, COVID-19 has forced ally affect how companies transform and
and the efficiency of operations through- every industry to reexamine its work pro- streamline their operations, and our cus-
out the asset’s lifecycle. Through 3D vi- cesses and digital twins offer a framework tomers are working with us to lead them
sualization, engineering data can tell cus- to reimagine the new world of work. through this period of change.
ON DEMAND WEBCAST
Ron Beck
Director, Marketing Strategy Managing Safety and Sustainability in a New Economic Reality
Aspen Technology, Inc.
The global pandemic has disrupted the process industries as never before. Shifting market
demands and an increasingly remote workforce are challenging companies to operate
safely, efficiently and profitably. At the same time, they cannot lose sight of their long-term
sustainability goals. Fortunately, advances in technology are helping downstream, upstream and
chemical leaders not just adapt, but find new ways to improve operations, reduce risk and create
value—all while driving toward a safer, more sustainable business.
Join AspenTech experts for a timely, informative webinar as they discuss how asset owners and
Ajay Lakshmanan operators can leverage digital innovations to:
Senior Director, Product Management
Aspen Technology, Inc. • Improve visibility across the organization to make faster, smarter decisions
• Empower remote users with the tools to optimize operations from anywhere
• Identify issues well in advance to ensure asset integrity, maximize uptime
• Use data analytics to identify improvements and reduce environmental impact
Discover how breakthrough technologies such as digital twins, AI and prescriptive maintenance
can be used strategically to sustain operations in any market condition.
20 FEBRUARY 2021 | HydrocarbonProcessing.com
Valves, Pumps and Turbomachinery
ometry of the internal passages. As such, as demonstrated in FIG. 2. PBF methods was cast iron and suffered from failure
the external geometry of the impeller was use either a laser or an electron beam to due to corrosion, the 3D printed impel-
obtained by scanning a new impeller that melt and fuse material powder together. ler was manufactured from 316 stainless
was available at a Saudi Aramco
warehouse. The internal geom-
etry was obtained by scanning
an older, damaged impeller after
Although the nonmetallic additive manufacturing
machining away its front shroud. industry has progressed significantly, metallic additive
Finally, a 3D model was de-
veloped for the impeller based manufacturing for the oil and gas industry is still in
on the two 3D scans. A 2D draw- the early development phase. Various technologies
ing was also developed for iden-
tifying the required machining have been developed, or are being developed,
tolerances and dimensions. The
3D model and the two drawings to allow manufacturing of metallic spare parts.
were shared with the Siemens
Energy AM center to manufac-
ture the impellers for this trial. FIG. 1 sum- The use of laser-based PBF technology steel material. This upgraded material
marizes the 3D model development pro- has various other names, such as selective was considered acceptable since the main
cess and shows the 2D drawing. laser melting (SLM), direct metal laser risk, which is galvanic corrosion, was
Note: The 3D model was modified by melting (DMLM) and laser powder bed found to be of no concern due to the low
the AM center to add extra thicknesses fusion (LPBF). total dissolved solids in the process water.
to all machined surfaces. This included The PBF technology is the most ma- As part of the manufacturing pro-
filling the keyway and adding material ture AM technology available for manu- cess, the Siemens AM center studied and
on the impeller wear rings landing areas, facturing complete parts. This technol- simulated the buildup process of the im-
bore and outer diameter. ogy is typically limited in size based on peller to minimize the needed external
the printer bed size, which is typically less supports and to eliminate any support
Impeller manufacturing. Since the se- than 400 mm × 400 mm × 400 mm. This inside the impeller internal passages, as
lected pump is not in a hazardous applica- size continues to expand, as some print- shown in FIG. 3. This is one of the most
tion and has a standby spare pump, the sup- ers have a higher build volume than this challenging manufacturing processes, re-
ply scope was limited to the manufacturing typical range. In addition, this technology quiring significant experience and tech-
of one impeller with test coupons on the has some limitations in terms of mate- nological know-how. It should be noted
same build bed. If the part was very critical rial. Typically used materials are copper, that the same model was shared with
or if the part failure presented a safety risk, aluminum, 300 series stainless steel, tool several other AM part manufacturers that
then the scope would have included the steel, cobalt chrome, titanium and In- rejected the work due to their inability to
manufacturing of one additional impeller conel. A market search shows no or very eliminate internal supports. The main is-
to be used for destructive testing. limited development on the use of this sue when adding internal supports is the
The AM technology used for manu- technology for manufacturing parts from inability to machine them away due to
facturing this impeller is the laser-based cast iron, carbon steel or duplex stainless limited accessibility, thereby rendering
powder bed fusion (PBF) technology. steel. As the existing impeller material the manufactured impeller useless.
Powder bed fusion is an additive manu-
facturing process in which thermal energy
selectively fuses regions of a powder bed,
FIG. 3. Printing arrangement for the impeller FIG. 4. Manufactured impeller after printing (time = 64 hr) (A); and after post-processing
and the test coupons. [(no heat treatment (B)].
The impeller, shown in FIG. 4, was the desired material properties supply chains and optimize investment
printed at a Siemens AM facility. Test without manufacturing defects in spare parts.
coupons were printed together with the 3. The qualification of the facility
impeller, using the same technique and quality system (e.g., by meeting ABDULLAH AL-GHAMDI is an
Engineering Consultant with the
parameters for the examination of mate- ISO 9001). rotating equipment division of the
rial structure and mechanical properties On the other hand, the qualification consulting services department of
(e.g., corrosion, tensile strength). The ac- process for an AM part can have signifi- Saudi Aramco. He has worked with
tual impeller printing duration was 64 hr. cant variation depending on the part criti- the company for 28 yr, starting as a
mechanical engineer with Saudi
The impeller was then removed from the cality. It can involve destructive testing of Arabian Marketing and Refining Co. (SAMAREC) in
printer for post-processing, which includ- manufactured parts, manufacturing and 1992 and joining Saudi Aramco in 1994. He served as
ed the removal of supports and final ma- testing of test coupons produced with the Standards Committee Chairman for pumps, seals
and mixers for 5 yr, as well as Responsible Standard
chining. The Siemens AM manufacturing AM parts, nondestructive testing for the Agent (RSA) for 3 yr. Mr. Al-Ghamdi received a BS
process for 316 stainless steel did not re- parts or mere visual inspection of the part. degree in mechanical engineering from King Fahd
quire any post-printing heat treatment. University of Petroleum and Minerals in Saudi Arabia
This trial showed the ability of AM to Part testing. The trial included nonde- and an MS degree in mechanical engineering and
dynamic systems from the Florida Institute of
significantly reduce the manufacturing structive testing for the impeller to verify Technology in Florida, U.S.
time for parts. The entire new part assess- its integrity through several tests, includ-
ment and manufacturing can be reduced ing visual inspection, radiographic test- SAAD H. AL-DOSSARY is the
Leader of the pumps group at
to 4 wk–6 wk, while printing of a previ- ing and penetrate testing. In addition, the consulting services department
ously completed part can be managed in a the manufactured testing coupons were of Saudi Aramco. He has worked
few days. It should be noted that the cost sent for material and tensile strength at Saudi Aramco for 20 yr. Mr.
of an AM part tends to be higher than the testing. The test coupons were printed Al-Dossary holds an MS degree in
the design of rotating equipment
original equipment manufacturer cost. in both horizontal and vertical arrange- from Cranfield University in the UK.
This is why AM should be considered ments, as shown in FIG. 3. This is typical
when the part is needed in urgent cases, for PBF technology, since the printing AMER AL-DHAFIRI is a Pump
Specialist at the consulting services
or if the end user is interested in reducing methods build up the part in the vertical department of Saudi Aramco.
its warehouse spare parts inventory. direction, which leads to some differenc- He has worked with Saudi Aramco
es in the mechanical properties between for 20 yr. He holds a BS degree
AM technology and part qualifi- horizontal and vertical testing coupons. in mechanical engineering from
King Fahd University of Petroleum
cation. The qualification for any part After passing nondestructive testing, and Minerals in Saudi Arabia and an MS degree from
manufactured by AM typically involves the impeller was installed in the pump the university of Virginia in the U.S.
the qualification of the AM production and went through a 5-mos field test where
MAJED AL-ZAHRANI is an
facility and the qualification of the part. the pump was kept in continuous op- Electrical Engineer at Saudi
At present, no international standards ex- eration. This field test included a pump Aramco’s Riyadh refinery with
ist that provide mandatory requirements performance evaluation that showed the 13 yr of experience in plant
technical and project tasks.
for AM qualification. However, several pump performed per its original design His work experience is focused
industrial standards have been devel- and within acceptable vibration limits. on plant technical issues; project
oped, or are in development, to provide Finally, the impeller was removed for in- design reviews; compliance assurance for standards,
good guidance on how to qualify an AM spection after completing the test. Visual procedures and material specifications; site
construction and commissioning; operations
production facility and an AM part. Ex- inspection and liquid penetrant testing support; and troubleshooting and fault investigation.
amples of these standards are DNVGL- were conducted, and the impeller was Mr. Al-Zahrani holds a BS degree in electrical
ST-B203, which address laser-based found to be in excellent condition. The engineering from King Abdulaziz University in
powder bed fusion (BPF-LB) technol- team decided to return the impeller to Jeddah, Saudi Arabia.
ogy and wire arc additive manufacturing the pump for permanent plant use. VLADIMIR NAVROTSKY is
(WAAM) technology. Chief Technology Officer (CTO)
The AM production facility qualifi- Takeaway. This article documents the of Industrial Application Service
for Siemens Energy and a Senior
cation typically consists of three main successful utilization of AM to manu- Principal Key Expert in gas turbine
sections: facture a pump impeller. This installa- design and aftermarket. He holds
1. Qualification of the printer, tion was part of the work being done a PhD in mechanical engineering
from Moscow Physical and Technical University.
which includes the qualification by Saudi Aramco to obtain experience Dr. Navrotsky has more than 30 yr of professional
of the printing machine and with this game-changing technology in experience in international organizations such as
associated auxiliaries, as well the oil and gas industry. With 3D print- the Central Institute of Aviation Motors in Russia,
as the qualification of the ing, companies can benefit from a sig- ABB in Sweden, ALSTOM in Switzerland and
Siemens. As an employee of Siemens for more than
printer operators nificant reduction in manufacturing time 20 yr, he has held several management positions
2. Qualification of the printing and consequently improve critical parts and was appointed Gas Turbine Fleet Director and
process, which typically includes delivery. This technology also presents Global Service Technology and Innovation Manager
significant development and a great opportunity to transform from (CTO) in 2007. In 2010, he was also promoted to
Senior Principal Key Expert in addition to these
testing to find the correct printer physical inventory in warehouses to a responsibilities. In 2015, Dr. Navrotsky was awarded
parameter settings that produce digitalized warehouse that can simplify the honor of Siemens Top Innovator.
22 FEBRUARY 2021 | HydrocarbonProcessing.com
Executive CORIE ALLEMAND, GLOBAL LEADER FOR OIL AND GAS
Viewpoint Stratus Technologies
Photo: Combining IIoT data and AI-infused analytics, AVEVA’s augmented, virtual
and mixed-reality software enables industrial users to connect and interact with the
digital twin of their facilities. Photo courtesy of AVEVA.
Special Focus Digital Transformation
C. LAU, Schneider Electric, Houston, Texas
makes sense to integrate power and auto- Microgrid optimization. Microgrid op-
mation that are traditionally separated. By timization is optimized energy manage-
adopting this unifying concept through ment from multiple distributed energy
the entire lifecycle of a plant—beginning sources, interconnected loads, controlled
with incorporating power data into the as a single entity, operating in parallel
digital twin from design to build to opera- with the grid or in an intentional island
tions to maintenance—profitability and mode. Hospitals and airports are utilizing
sustainability can be improved dramatical- microgrid to improve their resiliency and
ly. By integrating energy data with process energy efficiency. The microgrid strategy
data into a unifying process automation is not a new concept to oil and gas compa-
system, operators can visualize asset health nies. Plants have been optimizing energy
and energy efficiency; conduct rapid di- sources and trade-offs using steam, fuel
agnostics of motors, pumps, fans, MOV and electrical grid energies. Refineries
and electrical assets problems through se- have been utilizing cogeneration technol-
quence of events; diagnose problems; and ogy for a long time, while some plants
resume productions in minutes vs. days. are selling power back to the grid, and
most plants have fast load-shedding capa-
Energy sourcing. Strategically sourc- bilities. What will be new for oil and gas
ing energy can reduce energy cost. For companies is incorporating renewable en-
For more than a century, example, the author’s company can help ergies as part of the energy management
Cook Compression has customers make smarter purchases, system. Some onshore well pads and off-
developed innovative whether it is improving efficiencies or in- shore platforms are already incorporat-
technologies to move creasing renewable energy content. The ing solar and offshore wind energies. We
the compressor industry company has helped more than 6,000 cli- may see microgrid applications for down-
forward. Today, we are ents globally and managed $30 B in port- stream oil and gas soon.
folio spending to date.
ready for the next big
Energy as a service. This is a new con-
move – toward a carbon- Active energy management. Tradi- cept of operating companies outsourcing
neutral future. tionally, companies manage the three energy to a third-party company to design,
elements of energy efficiency, sourcing build, own, operate and maintain with no
With our brand-new and sustainability in silos (FIG. 1). A bet- capital investment. One of the major ben-
Innovation Lab, expert ter way of approaching this is by adopting efits is the shifting of CAPEX to OPEX,
engineers and dynamic strategic comprehensive planning to help especially when capital budgets are con-
TruTech™ materials reduce energy cost and lower emissions. strained. The idea behind this is that com-
The author’s company’s energy experts panies can choose to focus on their core
portfolio, we continue to
can help improve energy forecasting and business in oil and gas by leasing energy.
design new solutions and purchasing decisions, and benchmarking
set new standards for energy KPI goals. An energy manage- NOTES
performance. ment softwarea that is based on a global a
EcoStruxure™ Resource Advisor
data management platform collects vast
CONSTANTINE LAU is a Global
cookcompression.com/H2 data sources and integrates them into one Director of the Oil and Gas segment
single source of energy and sustainability with Schneider Electric, and has
information depository. more than 20 yr of experience in
industrial automation and software.
He is responsible for oil and gas
Value chain optimization. Enterprise- segment business strategies,
wide value chain optimization is based sales and marketing programs. Mr. Lau has published
on enterprise visualization and unified numerous articles on advanced software topics,
such as advanced process control and optimization,
a company planning and scheduling. By optimiz- AI systems and Fieldbus. He graduated with dual
ing feedstock purchases and production engineering degrees from the University of Texas,
schedules, companies can make faster and an MBA from the University of Houston, Texas.
26 FEBRUARY 2021
Special Focus Digital Transformation
M. HASTINGS, Brüel & Kjær Vibro A/S, Nærum, Denmark
tive, but insufficient from a maintenance for specialized monitoring and Boolean resolution that far exceeded that required
point of view. The data historian concept logic for generating alarms. for process data. An even greater obstacle
consequently emerged to fill this gap. With For all practical purposes, the data his- for the migration of CMS data into the his-
advanced data processing and storage ca- torian seemed like the perfect solution torian was the question of how to manage
pability, the data historian extended DCS/ match for integrating a CMS. Both systems vibration diagnostic expertise.
PLC capability by offering much more depend on big data and use much of the Much of the work in a data historian is
insight into production reliability, quality same information processing tools. The simply to detect and trend measurement
and performance for the entire plant, as integration, however, has been slow. Why? amplitude changes to identify process
well as for the individual machines. changes or developing machine faults.
Moreover, the historian provided Machine condition monitoring in A CMS also does this, but a major func-
tools that can work extensively with the the data historian. In the beginning, tion of the CMS and the specialists who
data. These tools are continuously being the data historian data was limited to the work with the system is to also identify
developed but presently include func- DCS/PLC plant-wide process control and localize the detected fault, determine
tionality for historical and current ma- data, whereas the CMS stand-alone sys- the nature and severity of it, and estimate
chine performance visualization, naviga- tems were focused on machine vibration lead-time to service. This is fault diagnos-
tion for finding data, event handling of data. This was initially a technology-based tics, a specialist area for vibration analysts,
anomalies, notification of events, report- decision since raw vibration data and spec- where most who work in the data histo-
ing, and even calculated measurements tral data required a data sampling rate and rian space are traditionally not trained in.
Although many vibration measurements,
such as the bandpass for running speed
(1x) and the second and third harmon-
ics (2x, 3x), can easily be stored, detected
and trended in the historian. Expertise is
needed to evaluate the relationship of this
data to perform proper diagnostics.
A CMS specialist (inhouse or from the
CMS supplier) is needed to look at and
diagnose the data stored in the historian.
Moreover, the CMS specialist utilizes di-
agnostic tools in the system that are not
normally available in the historian, as
shown in FIG. 1. Therefore, even if many
vibration scalar measurements can be
stored, detected and trended in the his-
torian, most of the diagnostic work is ac-
tually done in the CMS itself. Therefore,
FIG. 2. An example of a data historian view screen showing the health status for a compressor many have questioned the actual benefits
train. The circled vibration signal in the upper left-hand corner is a link to the diagnostic tools of storing the vibration scalar values in the
shown in FIG. 1.
historian rather than in the CMS server.
Because of the diagnostic expertise
barrier, condition monitoring functional-
ities that could otherwise be done in the
historian remain in the traditional CMS,
namely data storage, visualization, navi-
gation, event handling, notification and
reporting. This makes the CMS an iso-
lated world of its own. Is any work being
done to change this?
However, for the diagnostic functional- Now, diagnostic services are not only knowledge base, and more stakeholders
ity part of the system, such as diagnostic more readily available, but they can be who make important asset reliability deci-
algorithms, specialized measurements performed more reliably and accurately sions can have access to this information.
and plots, this is still separated from the (FIG. 3). The vast amount of process and
historian. However, even this has been machine monitoring data in the historian Takeaway. The IIoT has been the main
improved. Diagnostic plot functions can now be analyzed by statistical analy- driver for successfully breaking down
and their corresponding measurements sis to further refine symptom detection proprietary interfaces in the informa-
(FIG. 1) can now be accessed directly from earlier and provide more reliable progno- tion world for a large part of the indus-
the historian user interface with a single sis to service with greater lead-time. CMS try, and the machine condition monitor-
mouse click, as shown in FIG. 2. suppliers are now also offering automatic ing domain is no exception. As a result,
One of the most important advances, diagnostic services through a cloud- historian-base condition monitoring has
however, is that it is now possible to store based webserver, so no client diagnostic become much more intuitive, reliable,
high-resolution vibration time wave- server hardware or software is required. transparent and widespread. A lot of work
forms in the historian database. This was This cloud can then be connected to the remains to optimize the historian-based
the initial stumbling block for CMS-data historian as a seamless interface. condition monitoring solution, but this is
historian integration; it has finally been More and more historian-based con- expected to be refined over time.
removed so all data can now be stored in dition monitoring applications are being
the historian, both scalar and dynamic developed, such as specific monitoring MIKE HASTINGS is a Senior
values. This means raw data in the his- techniques like thermodynamic perfor- Application Engineer at B&K Vibro
torian can be re-processed by service mance monitoring of turbomachinery, or based in the Denmark office.
providers either as a second opinion, or for specific machines, such as reciprocat- He has been with the company
for 30 yr and has worked with
in the case where the normal diagnostic ing compressors. developing machine monitoring
specialists are unavailable. There is still Another important area of improve- strategies and optimization
no standardized way of associating mea- ment with historian-based condition techniques for a wide range of machines. He now
works with strategic market development, analysis
surement configuration information to monitoring solution integration is the and communications. He is also active in ISO
the raw data stored in the historian, but expanded user base. More specialists can committees that create standards for condition
this is expected to change, as well. contribute to the machine healthcare monitoring and diagnostics of machines.
LIVE WEBCAST
Tuesday, February 9, 2021 | 10 a.m. CST / 4 p.m. UTC
Join the webinar to learn best practices for ensuring the prevention of off-
spec products, with emphasis on products coming from amine treaters and
caustic treaters, as well as, steps you can take to bring off-spec product back
to appropriate quality should you encounter such an event.
BE PART OF THE
HYDROGEN REVOLUTION
H2Tech is a new initiative to better serve the hydrogen sector. Through
a quarterly technology journal, weekly e-newsletter, podcast and virtual
conference, H2Tech is devoted to applications and trends for the
hydrogen community worldwide!
H2TECH INCLUDES:
» Technology advances for » Safety and sustainability
hydrogen production and use » Market pricing trends
» Blue and green hydrogen projects » The future of hydrogen energy
» Hydrogen distribution and fueling » And much more!
infrastructure development
H2-TECH.COM #H2TECH
Special Focus Digital Transformation
A. BUENEMANN, Seeq Corp.,
Seattle, Washington
host of insufficiencies in their traditional Limited higher-order modeling func- effort ultimately results in reports full
spreadsheet-based calculation tools. tionalities require implementation of vi- of static images that require reconstruc-
Leading limitations include data connec- sual basics for application scripts or data tion any time a report is created for a
tivity and access, computational perfor- dumps into other advanced program- new time range.
mance, collaboration, versioning, visual- ming tools. Both options present addi-
ization and reporting functionalities. tional challenges regarding knowledge Advanced analytics solutions to
When establishing the root cause of a transfer and software tool maintenance. overcome limitations. Browser-
shutdown or a slowdown, it can be help- Multi-user collaboration has become based advanced analytics software ap-
ful to overlay process data with contextual increasingly important as workforces plications have filled the gaps left by
information, such as operator logbook adapt to a remote environment. Online spreadsheet-based calculations and
notes or maintenance work orders. Estab- tools allowing simultaneous input from reporting tools. Data connectivity and
lishing a live connection to these types of multiple users are replacing offline ver- refreshing, along with computational
data sources is challenging because, with- sions. Spreadsheets and traditional desk- load, interactive visuals, and live updat-
out it, engineers must extract data from top-based statistical modeling software ing summary dashboards and reports
multiple databases and build separate present version management and other are now well within reach.
queries for each. issues, creating a barrier to collaboration. Advanced analytics applications
Spreadsheets encounter performance For presentation or sharing, spread- provide out-of-the box connectors to
limitations when dealing with highly com- sheet-generated data visualizations are process data historians and SQL-based
plex calculations or high data volumes. The often embedded in slide decks or other contextual databases—making process,
latter of these becomes a serious problem reporting tools. These visualizations maintenance, shift log and other data
when performing historical data aggrega- can be tedious to construct due to chart available to SMEs from a single applica-
tions, as the volume of data compounds axes limits, as displaying more than two tion. Refreshed data is indexed on de-
with time. Spreadsheets reach row and col- parameters together requires manual mand in a live updating environment,
umn limits, and they become increasingly scaling, which is often specific to a sin- while maintaining the integrity of the
difficult to manipulate as file size increases. gle iteration of the visualization. All this data source.
Cloud or on-premises, server-based
computing is leveraged, empowering
SMEs to do calculations beyond the
capabilities of their respective singular
machines. This application saves infor-
mation after every click, thus alleviating
the threat of catastrophic crashes of large
spreadsheets hours after the most recent
save (an all-too-common occurrence
when dealing with large volumes of data
and complex calculations).
The combination of a versatile vi-
sualization pane and a robust calcula-
tion engine enables intermediate visual
feedback and rapid iteration of analyses.
Visualizing all the steps in a workflow
enables users to identify potential issues
sooner, troubleshoot more efficiently
and minimize recycle time.
A “set it and forget it” dashboarding
and reporting tool with a live connection
to the data embedded in the visualiza-
tions saves engineers significant time
that was previously spent updating and
formatting reports. Auto updating or
scheduled date range configuration en-
sures that high-horsepower calculations
are run at optimal times, with reports
made available for other users at a time
of their choosing.
have been deployed across various indus- These summary visualizations shutdowns and to create a golden
tries to enable customers to maximize are compiled into an organizerb, profile of these run cycles. The
long-term production, optimize mainte- where configurable live or forecasted profile was then used
nance spend and improve reliability—as scheduled date ranges are used to compare against the actual
shown in the following examples: to create automatically generated production rate to understand
1. Identification, categorization, periodic reports, complete where the projected end-of-run
summary and reporting of with collaboration and sharing date stood compared to the
production losses capabilities (FIG. 1). best-case order fulfillment date
Challenge: Process Result: Automatically generated (FIG. 2).
manufacturers need to track and monthly reports can save Result: A sold-out production
categorize performance losses 1 d/mos–5 d/mos of valuable unit had been looking at ways
to identify bad actors, justify process engineering time. This is to increase capacity to meet
improvement projects, and time that engineers can get back demand. By implementing this
perform historical and global to work on improvement projects proactive downtime strategy, they
benchmarking. Production and other value-added activities. were able to meet supply chain
loss accounting for process Easily exportable historical targets on average 11% sooner
manufacturing is a tedious loss data enables engineers over the course of the year. This
activity that can cost days of work to spend more time adding allowed them to creep production
for process engineers. It requires value to improvement projects volumes for multiple products
identifying losses, performing and less time developing by a proportional amount, thus
root cause investigations, and cost justifications. growing sales and market share.
documenting the events leading 2. Production run length
up to the loss. To make this optimization Takeaway. Reporting, complex calcula-
information available to all Challenge: Many process tions, modeling and other tasks required
interested personnel, it needs manufacturing units hit process for analysis of time series process data
to be easily aggregated into throughput constraints over the have traditionally been performed us-
reports that convey overall course of a run, resulting in the ing spreadsheets. As data volumes grow,
equipment effectiveness and degradation of production rates along with pressures to increase person-
identify unreliable equipment. over time. These constraints nel productivity, new solutions are need-
Effective use of this data can are often reversible, but at the ed to deal with these and other issues.
inform decision making on capital cost of shutting down to clean/ Advanced analytics applications that
project spending to remove maintain equipment. Sometimes, are specifically designed to deal with
process bottlenecks and upgrade the decision to shut down and time-series process data provide a solu-
unreliable equipment, with the clean equipment—regaining tion for large data volumes. As these use
end goal of minimizing downtime, the throughput rate upon cases demonstrate, utilizing the right tool
maximizing production rates startup—enables a unit to meet for the job saves hours or days of process
and optimizing maintenance. its production goals sooner. engineer time, freeing these valuable indi-
This categorization is hard to do This type of planned downtime viduals to work on higher-value activities
retroactively, and most accurate also reduces maintenance that will optimize production, increase
performance loss coding is done spend, as opposed to spending reliability and optimize maintenance.
either programmatically or by time reacting to equipment
frontline personnel at the time failure. Meeting targets sooner NOTES
of, or shortly after, a performance translates into more production a
Seeq Workbench
b
Seeq Organizer Topic
loss event. and increased profits over the c
Seeq Formula
Solution: Digital toolsa are used long term. Developing solutions
to identify performance losses to these types of optimization ALLISON BUENEMANN is a Senior
by comparing an operation to problems typically requires Analytics Engineer for Seeq Corp.
ideal situations and creating complex calculus, along with She has a process engineering
background, and spent her
conditions when the operation advanced modeling packages and early career with ExxonMobil,
is constrained compared to programming experience. working in petrochemical
target. Losses are categorized Solution: A calculation enginec and polymers manufacturing.
using specific conditions, where was used to calculate the number Ms. Buenemann has spent much of her time
completing improvement projects to debottleneck
events can be placed in the of shutdowns that would constrained unit operations and making operational
correct categorical buckets, either minimize the total time required decisions with the goal of maximizing production.
manually or logically, based on to produce a given order size. In her current role, she draws on that manufacturing
configured thresholds. Summary After determining the optimal experience to provide advanced analytics solutions
to customers in all process industry verticals.
visualizations are created in the number of shutdown/run cycles, She earned a BS degree in chemical engineering
application to represent the losses engineers were able to determine from Purdue University in Indiana, and an MBA
both graphically and tabularly. the run length between degree from Louisiana State University.
the minutes-to-hours effect of a change in the key variable on The model must also have the internal complexity required
economically critical temperatures, pressures, flows and lev- to capture subtle to severe nonlinear dynamics between its
els at other parts of the plant. Even if these simplified models inputs and outputs.
were perfect, the required feedstock information for the cal- Critically, the new model must compensate for missing
culation of the optimal variable value simply does not exist. feedstock composition data, even when that composition
The simplified high-level models can only provide rough, sub- varies subtly on a daily and hourly basis. The interconnect-
optimal, steady-state guidance to help drive the key variables. ing model must deduce this compositional information from
An unfortunate effect of the vast resolution difference be- other measurable real-time sources of information, such as
tween the lower layers and upper layers of the stack is siloed product yields and unit conditions.
teams. Process control teams are centered around the PID and This new closed-loop interconnecting model must not be
APC models. Planners and economists are concentrated on limited to single layers of the traditional stack. A single model
the full-plant LP model or scheduling model. Each modeling should be able to simultaneously operate at various layers, in
layer requires wildly different skills, experience, terminology different units and areas of the plant, and to interface differ-
and technologies. As a result, these teams often have their own ent teams and engineering disciplines. It must understand the
goals and initiatives and are not well positioned to collaborate global nature of the optimization problem at hand, without
in optimizing the plant. losing the detailed resolution for local process subtleties.
The traditional hierarchical stack can solve simple multi- To capture the lost margin, refiners and chemical operators
unit coordination and optimization problems, such as prod- are looking to artificial intelligence (AI) and machine learning
uct blending. These are special cases where the relationship (ML) to implement this new approach. Unfortunately, simply
between key process parameters and the resulting economic adding AI and ML capabilities into the stack will not provide
properties is linear or easily linearized. It is only the large num- the desired step change and will not capture the lost margin.
ber of settings to be manipulated simultaneously that makes
product blending an optimization opportunity. For such cases, Integrating AI into the traditional layered stack of
a simplified subset of the plant LP model can run on a minute- process control and optimization. Introducing AI and ML
wise or hour-wise frequency to coordinate the units and cap- capabilities into the current stack may improve engineering
ture most of the value. productivity within each layer. Process control teams might
When it comes to more complicated (and more profitable) be able to tune PID controllers faster and maintain APC mod-
parts of the plant, the hierarchical approach cannot capture els more efficiently with less invasive step testing. Planning
the full optimization opportunity. Conversion units are an ex- and economics engineers may get better tools to consolidate
ample. First principle process models or simplified LP models multiple large spreadsheets. These are the types of advance-
can generate a rough recommendation for reactor temperature, ments one might expect from augmenting the current layers
accounting for feedstock and product economics. However, with AI and ML.
conversion unit reactors exhibit substantial nonlinear dynamics A nascent industry discussion is the combination of first
and are sensitive to the slightest changes in feed composition. principle models with AI and ML for closed-loop optimiza-
Even 3°F–5°F away from the optimal reactor temperature im- tion. However, the use of statistics in this respect is not new.
plies over-cracking or under-cracking the feed and is detrimen- For the several decades in which first principle models have
tal to product yields. In these cases, the first principle model been in use, plant data has always been employed to reconcile
does not operate with enough resolution to manipulate the and fit the models to the current state of the plant. In fact,
reactor temperature on a minute-by-minute basis, with respect the recent scientific breakthroughs and developments that
to disturbances and ever-changing feed composition. Since feed are driving AI and ML into public awareness are the result
composition is not reliably measured in real time, it is literally of going beyond first principle models. Recent AI discover-
impossible to solve this problem via the traditional stack. ies are focused around enabling algorithms to build their own
process representations and models.
An interconnecting closed-loop model. Capturing the For example, one may consider using a combination of ML
lost margin opportunity of closed-loop optimal manipulation and nonlinear dynamic first principle models for closed-loop
of the plant’s key levers requires refiners and chemical opera- optimization of complicated conversion units. In this case,
tors to break the mold. It requires them to look beyond the thousands of parameters must be calibrated in real time by us-
traditional hierarchical stack and to build a closed-loop inter- ing some form of ML. There may not be enough process data
connection between planning and economics, process engi- to automatically fit the large number of model parameters in
neering, process control and operations. real time. If the first principle model is simplified enough to al-
This new interconnecting model is not meant to replace the low automated ML-based fitting, it will not capture the subtle
traditional stack. In fact, for most of the control loops, the re- nonlinearities that the critical valuable opportunities require.
placement of linear technologies by a new approach will not Furthermore, first principle models require feed composition
provide a meaningful return on investment. The new intercon- data, which is not measured adequately and is, therefore, not
nection model must be tightly built around the set of 10–15 available as a reliable or accurate input.
most economically critical process parameters. Decades-old statistical model-fitting methods are now
As opposed to augmenting a single-layer first principle op- being rebranded as ML or AI. While some of these methods
timization, the interconnecting model must be able to truly might locally improve engineering productivity, they do not
manipulate the critical process parameters in closed loop. provide step-change improvements in plant profitability.
Hydrocarbon Processing | FEBRUARY 2021 35
Digital Transformation
36 FEBRUARY 2021 | HydrocarbonProcessing.com
Special Focus Digital Transformation
S. SINGH, ABB, Oslo, Norway
The product lifecycle is concerned resulting from these behaviors must flow Driving global standards. Manufactur-
with the information flows and controls, up and down the pyramid. This integra- ing organizations seek a manufacturing
beginning at the early product design tion from machine to plant to enterprise architecture that will remain consistent
stage and continuing through to the prod- systems is vital, and it critically depends with a broader enterprise architecture for
uct’s end of life. The existing standards— upon standards. global visibility, collaboration and control,
particularly, in the areas of computer- Standards-enabled smart manufactur- while still being flexible enough to sup-
aided design (CAD) and computer-aided ing integration allows the following: port individual site goals. This is a chal-
manufacturing (CAM)—have greatly im- 1. Access to field and plant data for lenge, especially for those who have inher-
proved engineering efficiency. In addition, making quick decisions and for ited multiple divisions and sites, as well as
these standards enhance modeling accura- optimizing production throughput multiple manufacturing styles and models
cy and reduce product innovation cycles, and quality (i.e., in-house, virtual or contracted).
thus contributing directly to manufactur- 2. Accurate measures of energy and Global standardization of the end-
ing system agility and product quality. material use to-end supply network is done through
The production system lifecycle fo- 3. Improved shop floor safety common process standards that are trans-
cuses on the design, deployment, opera- and enhanced manufacturing parent and supported by clear key perfor-
tion and decommissioning of an entire sustainability. mance indicators (KPIs). These standards
production facility, including its systems. Generally, existing manufacturing stan- provide a foundation for organizational
Areas of standards that support produc- dards provide how-to instructions for de- consistency, without prohibiting lo-
tion lifecycle activities include produc- signers, engineers, operators and decision cal flexibility, reliability and innovation.
tion system modeling data and practices; makers to conduct disciplined activities While local execution is enabled by local-
production system engineering, opera- within their domains. They also facilitate izing global process standards to market
tions and maintenance; and production communication between stakeholders specific best practices that consider the
lifecycle management. across domain borders, borders of the varying asset structures, other consider-
The business cycle addresses the func- manufacturing system hierarchy and be- ations (such as cultures, regulatory com-
tions of supplier and customer interac- tween lifecycle phases. Standards and ref- pliance and other local factors) have his-
tions. Standards for interactions among erence models can offer an organization torically created constraints. These best
manufacturers, suppliers, customers, part- through a baseline for common lexicon practices are continuously improved to
ners and even competitors include general for consistent engagement across different meet business objectives that are continu-
business modeling standards and manu- functions and geographies. At the most ally being reshaped by market dynamics.
facturing-specific modeling standards and basic level, reference models and stan- Global process standards drive con-
corresponding message protocols. These dards will help with business cases, tech- sistency and reliability in manufacturing,
standards are the key to enhancing supply nical feasibilities and value proposition but deciding how far to extend them into
chain efficiency and manufacturing agility. evaluations. At a more detailed level, some local manufacturing operations without
Each of these dimensions comes into might lend process maps and templates compromising agility is hard. The balance
play in the vertical integration of ma- that help identify assets, applications and is tipping to a point where site autonomy
chines, plants and enterprise systems data, as well as potential resource alloca- and control over manufacturing systems
called “the manufacturing pyramid” tions and security requirements—all of is giving way to manufacturing standards
(the central pentagon in FIG. 2). In smart which are helpful for scale should proto- mandated by enterprise architects. At the
manufacturing, autonomous and intelli- types be proven. However, standards still heart of these debates is how far to take the
gent machine behaviors—including self- point toward individual processes and use standardization of processes, information,
awareness, reasoning and planning, and cases vs. a complete smart manufacturing technology and solutions, while allowing
self-correction—are key, but information concept. for specific local plant or functional needs.
FIG. 3 provides a process and system map
where organizations can define how far to
extend global standards into local manu-
facturing operations, and how to follow
local best practices without compromis-
ing reliability and performance. Enter-
prises can define the point (or “locus”)
where unified and global standards, along
with best practices and systems, align to
deliver value to the business.
Beyond this point, the lack of local
flexibility, as well as the time and cost to
establish standards, fails to deliver busi-
ness benefits. For example, a precision
parts producer that supplies piston pins
to multiple automotive OEMs worldwide
FIG. 2. Smart manufacturing standards (adapted from NIST IR 8107).
requires consistent quality. This producer
38 FEBRUARY 2021 | HydrocarbonProcessing.com
Digital Transformation
HYDROGEN
Global Energy Infrastructure (GEI) is a new market intelligence tool for the international
energy industry. With the recent disruptions across all energy markets, it’s more important
than ever to get up-to-date information on projects, including project status.
Combining our well-established downstream project data set, used by the biggest names
in global oil and gas, with new data sets for hydrogen, renewables, global pipelines, and
LNG, GEI gives you the intelligence you need to grow as the energy industry recovers and
transitions.
Designed with ease-of-use in mind, GEI equips you and your team with access to the
latest data in an easy-to-access and easy-to-use format. All data and market intelligence
reside on the GEI website, updated 24/7/365 by our researchers in Houston, London, and
Mumbai.
For customers who need the data in spatialized formats, Energy Web Atlas
continues to provide a superior tool which maps global energy assets for
midstream, downstream, renewables, and hydrogen.
is achieved by the enclosing of all op- coke and water are separated. Each bin is The design made for the Grupa LO-
erational steps in a “closed system.” FIG. 2 equipped with internal screens and can TOS Gdansk refinery was a grassroots
shows a schematic of the CCSS. hold one batch of coke. The majority of project and the first of its kind. All new
In the closed system, the coke and fines produced from the cutting proce- CCSS structures may be installed in the
water from the coke drum discharge to dure are trapped instantly in this coke pit/pad area, and integration and re-use
a special coke crusher via a transition bed, resulting in clean water filtrate. The of existing equipment can be investigated
piece. The transition piece is a permanent clean water filtrate is collected in the drain with a tailored customization. Several
and sealed connection between the coke water basins and routed to the water set- revamp studies, with concepts for maxi-
drum bottom unheading device and the tling and clean water tank for final clean- mum modularization and pre-fabrication,
coke crusher, capable of tolerating the ing and re-use for the coke cutting equip- have already been executed and are of
thermal expansion of the coke drum. The ment, slurry transport and quenching. high interest to all units considering the
inline coke crusher reduces the size of After completion of dewatering, the elimination of fine emissions.
the coke particles suitable for processing, coke is discharged from the dewatering
storing and selling the green coke prod- bin via a vibration feeder, directly onto a Startup challenges and trouble-
uct. FIG. 3 shows a picture of the transition belt conveyor, which transports the coke shooting. The Grupa LOTOS DCU,
piece and the coke crusher. product to the coke storage area. featuring the proprietary CCSS, was com-
Downstream of the coke crusher, a The integration of the closed system missioned in September 2019. Other than
slurry basin collects all incoming streams includes the intake of all streams, which some challenges with the hydraulic decok-
from the coke drum; water is added to the are normally routed towards the pit/pad. ing system, which had some impact, this
coke to form a pumpable slurry. A special Simultaneoulsy, the closed system pro- first-ever commercial demonstration of
slurry pump transports the water-coke vides clean water to the DCU for quench- the CCSS was an opportunity to detect
mixture to closed dewatering bins, where ing and cutting. any other unanticipated issues, which can
The main advantages of the CCSS vs. now be addressed in subsequent designs
For re-use conventional open systems are: and implementation. These issues—
Coke Settling and clean • The elimination of dust emissions which mostly involved plugging in parts of
drum Coke water tank to the atmosphere due to the the CCSS—and how they were handled
fines closed design are furthered detailed in the next section.
Transition
piece Dewatering bin
• Low workforce requirements Hydraulic decoking system. During
Crusher Coke due to automation the initial period, the coke cutting pump
Drain water
• Reduced water consumption tripped several times due to high bearing
Slurry basin • The reduction of equipment
basin temperature, which caused delays in the
Slurry pump Drain water pump footprint decoking time. These delays impacted the
• More flexibility in equipment coking/decoking schedule, and the unit
FIG. 2. Schematic of the CCSSb. arrangement. could not be operated at the unit design
throughput. The hydraulic decoking ven-
dor was contacted, and upon the replace-
ment of the shaft, impeller and drilling
nozzle, the coke cutting pump tripping
issue was resolved.
Plugging of the slurry pump and
slurry line. During commissioning, plug-
ging of the slurry pump and slurry line was
observed. An additional flushing connec-
tion was provided on the suction side of
the slurry pump, which helped avoid cavi-
tation and improved stability of the flow
pattern. In addition, the flow control logic
of transport water was simplified, which
contributed to a more stable operation.
Plugging and leaking of the transi-
tion piece. During initial operation, plug-
ging was observed in the transition piece,
which pushed some coke into the body of
the bottom unheading device. The plug-
ging forced a back-up of water and even-
tually caused a breakthrough, in which
a large amount of water and coke would
suddenly discharge from the coke drum.
FIG. 3. View of the transition piece and crusher. This caused the slurry basin to overflow
42 FEBRUARY 2021 | HydrocarbonProcessing.com
Process Optimization
and carry coke particles into the neigh- Plugging of feed line. During opera- watering cycle.
boring drain water basin. tion, plugging was observed in the feed After gaining some experience with
From the drain water basin, coke par- line. This was mitigated by modifying the dewatering rates, an improved operat-
ticles were pumped into the water set- the decoking permissive to allow water ing procedure was established, and coke
tling tank, which resulted in the blocking through the feed line during cutting. moisture reached an acceptable level.
of the desludging line. Moreover, some Instrumentation. In some events, the
sludge even carried over into the clean sudden filling of the slurry basin resulted Takeaway. The first implementation of
water tank. Because of this, routine mon- in the radar-type level measurement the proprietary CCSS was successfully in-
itoring of the sludge levels in the water sensor to be “dipped” in the liquid. The tegrated with a licensed DCU. Successful
tanks was implemented. This step helped radar sensor can also be affected by the execution in commissioning showcased
detect abnormal levels of sludge in the presence of steam. These events caused a path to improved sustainability that
water tanks and allowed for corrective ac- the measurement to freeze at its previous refiners can follow as an alternative to a
tion. The desludging line was unblocked, value for approximately 20 min each time conventional open pit/pad coke handling
and both water tanks were flushed. It was and resulted in the level control being in- process. The lessons learned during this
also recommended to install a coarse effective and taken to manual control. In- project have helped improve CCSS engi-
screen in the overflow between the basins stead of a radar sensor, a second differen- neering design practices. Future projects
to keep the coke particles inside the slur- tial pressure sensor can be implemented are expected to have more robust designs
ry basin. Future slurry basins will be de- to alleviate this issue. that present a much cleaner alternative to
signed to withstand a sudden discharge Dewatering/coke transport. During an open pit/pad coke handling system. All
or breakthrough of the coke drum. startup, the coke discharged to the trans- of the existing open pit/pad systems where
In addition, leakages were observed in port system was still wet and partly un- refiners wish to minimize coke fine emis-
the flanges of the transition piece. This suitable for inclined conveying. sions may become candidates for revamp
issue was resolved by adding hydraulic The quality of coke has an effect on to the environmentally friendly CCSS.
clamps to the flexible connection of the the cutting and dewatering times. Harder
transition piece, which are activated/ coke requires extended coke cutting, and NOTES
tightened during the water quench and powdery coke needs more time for dewa- a
Chevron Lummus Global
coke cutting steps. tering. Both effects result in a longer de- b
TRIPLAN’s Closed Coke Slurry System (CCSS)
NEW VERSION
InstruCalc •
for additional water or air coolers also resulted in lower vapor condensable Takeaway. In conclusion, optimization
for final heat rejection. This is and noncondensable flows at the top of of column operating conditions and
achievable through the use of a the column. These approaches avoid ma- design is critical when the column is al-
higher-efficiency bed at both the jor modifications to the VDU column, its ready operating at its rated capacity in
TPA and MPA sections, allowing auxiliary systems and the heat exchanger a brownfield project. A case study of a
better heat recovery at lower network, significantly reducing the capi- VDU column in a Malaysian oil refinery
pumparound flow, which reduces tal investment of the brownfield project. indicates that massive modifications are
both the column internals vapor required if optimization is not carried
and liquid traffic and improves Results and discussion. The modifica- out on the column operating conditions,
heat recovery in the highly tions required for the VDU column with which add unnecessary capital expendi-
integrated heat exchanger network. and without the optimization are listed in ture and may render the project econom-
5. For all fractionation packing beds TABLE 1. ically unviable.
within the VDU unit (i.e., the The optimization of the VDU column
fractionation and wash sections), operating conditions and internals design IR. SAIFUL ANUAR MOHD
hybrid packed beds were chosen avoids the replacement of the first-stage MOKHTAR has more than 18 yr
to provide better performance ejector (the largest ejector system), the of work experience in the refining,
natural gas/LNG and petrochemical
(i.e., a higher void fraction without associated steel structure for the ejector, sectors as an Engineering Manager
sacrificing efficiency). By using the requirement for a new cooling water and Head Process Engineer for
the high-efficiency packed bed, package and associated interconnected various Petronas capital projects.
At present, he is the Head Process Engineer
the hydraulic limitation was piping, a new steam generation package for Petronas Chemical Group’s titanium program.
reduced without having to increase and associated interconnected piping, a Mr. Mokhtar is a Chartered Chemical Engineer
the column diameter, while still new sour water treating unit, additional (CEng) with IChemE and a Professional Engineer
meeting separation and product heat exchangers, and cooling water or air (PEng) with the Board of Engineers Malaysia.
quality requirements. This choice coolers in the heat exchanger network. MARINA MOSIR has more than
avoids the need to modify the The total cost avoidance as a result of 15 yr of work experiences in the
VDU column shell. the optimization work is estimated at ap- refining and natural gas/LNG
industry as a Process Engineer,
With the considered optimization ap- proximately $25 MM, which allows the Lead Process Engineer and
proaches, it was observed that the VDU project to stay economically viable. Engineering Manager for various
column total internals vapor and liquid Additionally, the project recently un- Petronas operating units and
capital projects. She is a mass transfer design expert
traffic was greatly reduced throughout the derwent a performance test run and was within Petronas, with more than 12 yr of experience
column, while maintaining the products able to meet the desired performance in revamping, troubleshooting and designing
separation and quality. The optimization guarantees. distillation columns and separators.
46 FEBRUARY 2021 | HydrocarbonProcessing.com
Process
Optimization
K. MCINTUSH, J. P. FARONE,
B. D. PIGGOTT and C. M. BEITLER,
Trimeric Corporation, Buda, Texas
tion since most of the sour water in the fa- bustion device or fuel gas as allowed by The temperature of the sour water in the
cility has to be treated in the SWS before environmental regulations. At some sites, tank is usually less than in the flash drum,
it can be reused or processed further. The the flash gas is routed to the SWS over- which further reduces the solubility of hy-
sour water fed to the SWS will also change head gas line; however, this can result in a drocarbons in the sour water. Some in the
industry have also observed
that NH3 or amine-laden wa-
The SWS system is a critical unit operation and can be ter increases the solubility of
a significant bottleneck to facility production rates certain types of gasoline and
higher boiling range aromatic
if it is not adequately sized or forced to operate at hydrocarbons (similar to ben-
partial loads due to maintenance issues. zene) in the sour water, mak-
ing it difficult to separate such
that significant fouling was
over time, with increasing or decreasing significant and variable quantity of hydro- observed in the stripper.6 The sour water
amounts of NH3 and H2S in the water carbons being fed to the downstream unit surge tank may not remove all hydrocar-
and overall water flowrates varying daily. [e.g., a sulfur recovery unit (SRU) or other bons that remain in the sour water after
The designer of the SWS is challenged to technology] that can adversely impact per- the surge drum, but the tank does help by
design a flexible and robust system that formance of that downstream unit.7 Flash removing at least some of them.
can meet a variety of different feed con- gas with no condensable hydrocarbons The surge tank also allows for mixing of
ditions, while also minimizing the cost of could possibly be routed to the quench the sour water from different time periods,
the equipment. Above all, the SWS can- tower in an amine tail gas treating unit.7,8 so the composition is more uniform. If the
not be a bottleneck in the overall facility The sour water fed to the flash drum sour water composition changes consider-
and must strip the sour water reliably in often also contains liquid hydrocarbon/ ably or rapidly, the stripper may not func-
all operating conditions. oil that needs to be removed to protect the tion appropriately. By keeping the sour wa-
There is a long history of technical pa- rest of the SWS system from fouling and ter flow and feed composition consistent,
pers that thoroughly discuss many aspects prevent foaming in the stripping column. the stripper will be easier to control, and
of SWSs.1–6 This article is not meant to be The flash drum is usually a three-phase, a more consistent treated water product
a comprehensive review of SWSs. This horizontal vessel. A baffle system installed can be achieved. Short-circuiting—where
article reviews a few of the key design at one end of the flash drum is often used to the sour water entering the tank inlet flows
choices available for the SWS system, and skim oil from the water before it is pumped preferentially to the tank outlet without
then specifically focuses on some of the to the sour water surge tank. The oil over- adequate mixing or residence time—is a
internals of the SWS tower. The choice of flows the weir into a collection compart- common problem that results in higher
internals in the SWS can be difficult, with ment in the sour water flash drum for re- variability in sour water composition and
a range of different sources available in lit- moval. Another means of collecting oil is poor hydrocarbon separation. In the ques-
erature, and few very thorough technical to install a draw-off box in the sour water tion-and-answer portion of a recent indus-
analyses completed to guide the designer flash drum that could collect the oil over- try trade symposium9, measures that were
to the best solution. flowing to it. The minimum recommend- said to mitigate short circuiting included
ed residence time for the sour water inside (1) having the entry and exit on opposite
Auxiliary sour water separation the flash drum is 20 min, with a liquid level sides of the tank and (2) having the entry
equipment. For the SWS tower—and of 50%–60% being optimal. The sour wa- and exit at different heights.
the internals discussed later in this arti- ter flash drum should include connections The surge tank can be a fixed or float-
cle—to function properly, they must not for level bridles on the hydrocarbon and ing-roof-type storage tank. Floating roofs
foul too quickly. Sour water stripping is water side of the vessel. High- and low- can be either open or internal. However,
generally considered a severe fouling ser- level alarms and pressure indication are due to the potential for odors, a fixed-roof
vice. The stripper functions much better also used. Demisting equipment or other tank is often used. FIG. 2 shows an example
if the chances for fouling and foaming are similar plugging-prone internals are typi- of a surge tank with a fixed roof and internal
reduced by adequate pretreatment of the cally not used in the sour water flash drum floating roof. The floating roof may have a
sour water. Therefore, this section exam- because they may rapidly plug or corrode. double-seal design to minimize emissions.
ines the equipment upstream of the SWS The hydrocarbons collected in the sour Vacuum breakers and pressure relief
that reduces fouling and foaming issues in water flash drum are often pumped to a valves should be installed on fixed-roof
the stripper tower. slop system for further processing. tanks that are not vented to the atmo-
Sour water flash drum. As shown in Sour water surge/storage tank. The sphere. By letting air in, vacuum breakers
FIG. 1, sour water is collected in a flash drum sour water from the flash drum is fed to a can keep the tank from collapsing during
where hydrocarbon vapors and liquids are surge/storage tank. The tank is designed pump-out or upon cooling; however, air
removed. The vapors are flashed close to with several days of storage in case the ingress can lead to the formation of a dan-
ambient conditions to remove as much hy- SWS goes down. With long residence gerous combination of oxygen, hydrocar-
drocarbon as possible. The flashed gas is times, dissolved hydrocarbon liquid and bons and H2S in the tank headspace that
typically sent to a low-pressure destination emulsions can separate from the water and could lead to an explosion. Nitrogen or
such as a flare gas recovery system, com- collect at the interface level in the tank. inert gas blanketing is often used for this
48 FEBRUARY 2021 | HydrocarbonProcessing.com
Process Optimization
reason. However, inert blanketing has its Angled ports are sometimes installed Hydrocarbons in the stripper overhead
own problems. For example, using an in- on the sour water tank so that it can be gas can also cause operational issues in
ert blanket can lead to the formation of vigorously circulated (e.g., with a large the downstream SRU or other processing
pyrophoric iron sulfides on exposed steel portable pump) to stir up solids and then technology. Hydrocarbons in the stripper
surfaces. If air is subsequently allowed filter the solids out during turnarounds. bottoms that are routed to a water treat-
into the headspace (e.g., due to a fault in This reduces the frequency with which ment plant can pose environmental/regu-
the system or due to accident) and thus persons will have to go inside the tank to latory concerns, as well. Therefore, using
creates an explosive mixture, then the py- clean it out. Images of the usually unin- filters and liquid coalescers can benefit
rophoric material can ignite that mixture sulated exterior walls of the tank from a not only the SWS but also overall refinery
and cause an explosion. Examples of ex- thermal camera can sometimes be used operations.
plosions that have happened in SWS stor- to evaluate solids levels. In addition,
age tanks are documented in literature.10 the surge tank should have a bypass line SWS diameter and feed water feed
Much care is advised in designing inert around it so that it can be bypassed (e.g., locations. Many SWSs experience
blanketing systems for sour water tanks. for inspection), if needed. severe foaming, which needs to be ac-
The tank may need to have a minimum Sour water solids filtration/coalesc- counted for when sizing the column. As
of three days retention time during nor- ing filters. Additional solids filtration and such, the capacity should be de-rated to
mal operation at 50%–60% full, includ- coalescing technology may be installed account for foaming, and a system fac-
ing another couple of days of capacity for downstream of the sour water charge tor of 0.6–0.7 is typically recommended.
sour water storage. Whether the tank has pump and upstream of the feed/effluent This can make the SWS much wider in
a fixed or floating roof, it is common to exchanger. Solid particle filters should be diameter than would be anticipated for a
allow a hydrocarbon layer to float above used upstream of a liquid/liquid coalescer. column that, at least on first appearance,
the sour water as a “blanket” to limit va- Suspended solids removal improves the is basically boiling water.
pors from escaping that may be odorous efficiency of the coalescer by weakening The location of the sour water feed in
or toxic. This layer may be a diesel range the hydrocarbon emulsion and minimizes the stripper can vary based on several fac-
material and is sometimes also referred fouling from solids in the sour water heat tors, including whether trays or packing
to as a rag layer. Oil skims should be used exchangers, stripper reboiler (if used) and are used, number of trays used, the desire
to remove oil as the floating layer grows. stripper trays or packing. Some refineries for lower steam usage, inlet H2S and NH3
Floating skim nozzles with a non-me- reportedly have used a strainer instead of concentration and treatment specifica-
tallic flexible hose are sometimes used. a more expensive filter. tion, as well as operating temperature and
The design should include an automatic The liquid/liquid coalescer helps to pressure. If a pumparound cooling system
tank level gauge system, with provisions control hydrocarbon fouling in the same is used in lieu of an overhead condenser
for measuring the thickness of the hydro- sour water equipment. Disposable, micro- and reflux drum, the feed location will be
carbon rag layer on the aqueous layer, as fiber-based coalescers are reported to give below these trays, as well. Optimal feed
well. A literature source11 reports that nu- adequate separation of hydrocarbon emul- location can be determined in a process
clear signals or sound waves can be used sions.11 During a recent industry sympo- simulation, and the feed location is usu-
to measure the interface, but the authors’ sium9, use of liquid/liquid coalescers for ally located within the top several trays in
company is aware of successful measure- partial treatment of the feed sour water trayed columns. In addition, if the column
ment using capacitance probes, as well. was reported to increase the time between is constructed from carbon steel, it may
In a floating-roof tank, the capacitance cleaning from once every 9 mos to twice be lined with a corrosion-resistant dura-
probe can be mounted on the floating that length in a refinery with three SWSs. ble coating or made of corrosion-resistant
tank roof. Level control is critical to mini-
mize hydrocarbon carryover to the SWS; On loss of purge, alarm to control room Vent
the location of the control devices is vital N2
to accurately measure the interface.
The tank is typically made from car- Internal floating roof
bon steel, and a suitable durable coating
may be used on all interior surfaces to
minimize corrosion of the tank surfaces. Sour water surge/
Solids and heavy oils will sink to the Sour water from flash drum
storage tank
bottom of the tank. For this reason, the Sour water to stripper
tank bottom should be designed to slope
(e.g., approximately 3 in. for every 100
in.) to a low point drain. The tank dis- Sour water
charge pumps
charge to the pump is also generally el-
Slop oil to refinery
evated somewhat above the tank bottom
to allow room for heavy materials to ac-
cumulate without exiting the tank with Slop oil pump
the sour water. The sour water is pumped
FIG. 2. Simplified schematic of a sour water surge tank (internal floating roof).
using flow control to the stripper.
Hydrocarbon Processing | FEBRUARY 2021 49
Process Optimization
alloy above the liquid feed nozzle, where • All trays should be constructed of of thumb, but noted that actual tray effi-
corrosion is more significant. 300-series stainless steel or better. ciencies experienced in sour water service
Depending on the sour water (presumably well designed) were closer
Tray tower design for SWS. Most SWS processing demand, the tower to 50%. In designing a trayed system, one
systems are designed with trayed towers. may be too small for personnel could probably rely on the rule of thumb
Trays can be designed to be fouling re- to physically install the trays. to result in a system with significant over-
sistant. However, even in trayed systems, In this instance, cartridge trays design built in. For a less conservative and
the selection of an inappropriate tray can could be used. more economical design, careful engi-
lead to poor performance of the SWS. • If a pumparound system is installed, neering analysis and comparison with the
General recommendations for SWS tray the trays used for the pumparound actual performance of other similar SWS
selection include: loop should not be counted as systems is needed.
• Trays should be a fixed-valve type active mass transfer trays. Some factors that influence the effi-
and should be designed for vapor • In a fouling service like with an ciency of the trays include the following:
to flow horizontally out of the SWS, the downcomers are potential • Perhaps most importantly, tray
valves to minimize bridging of traps for fouling material and efficiency is a chemical engineering
deposits on the fixed valves. can adversely affect the capacity factor that is applied to equilibrium-
Tray designs like this are readily of a tray. Special designs that based designs to account for the
available from major distillation are available from the internals fact that operating trays do not
internals vendors. Sieve trays suppliers to address fouling material reach equilibrium conditions.
can also be fouling resistant in in the downcomers should be used. Hatcher and Weiland14 show that
some services. For example, the Tray efficiency is reported in several component efficiencies for H2S
authors know of acceptable sieve different ranges for SWS service, but gen- and NH3 will vary widely across
tray use in aqueous systems erally will vary from 15%–50% depending the stripper column, and could
with solid particles circulating on different factors. The number of trays depend heavily upon the stripped
(i.e., in slurry service). However, present in the SWS will then also vary water specification for the water
sieve trays have shown severe widely; a common range on the number of leaving the bottom of the stripper,
fouling in SWS service, with vapor actual trays installed may be 20–60. On a the steam rate to the stripper
flow area decreasing by as much 24-in. spacing, this translates to 40 ft–120 or reboiler, etc. Therefore, the
as 90%. This may be due to the ft of height for trays, which may mean an efficiency of the tray is not a static
vertical direction of the vapor SWS as tall as 150 ft in some applications. value throughout the stripper, varies
leaving the tray deck, which allows From the authors’ discussions with from one component to another
precipitation on the tray deck that a few refinery subject matter experts and may be different in the top of
can foul the tray.12,13 FIG. 3 shows (SMEs), a rough rule of thumb for de- the tower than it is in the bottom.
an example of fouling that can sign tray efficiency in an SWS is 3 actual To reduce uncertainty, the designer
occur on sieve trays in SWS service. trays per 1 theoretical stage or 33% effi- may need to do a more rigorous
This level of fouling occurred ciency. This is probably a conservatively simulation of the column.
over a typical SWS run between low efficiency for most systems. For ex- • The most important consideration
maintenance intervals of 5 mos.13 ample, one SME acknowledged this rule for SWSs is that they work reliably.
As a result, designs for SWSs tend
to be conservative. One way of
introducing conservatism into the
SWS design is to specify a low tray
efficiency that, when installed,
will allow the stripper to operate
and meet specifications in a more
heavily fouled state and to meet
specifications if the impurities
present in the sour water exceed
the initial design values. If there is
access to an existing stripper in the
same service, then operating data
can be obtained to verify the
design parameters.
• In many instances, the actual
composition of the sour water
feeding the SWS system may be
uncertain. Crude oil slates in a
refinery can change frequently, with
FIG. 3. Fouling of sieve tray in SWS service.13 FIG. 4. Example of tray fouling in SWS service.13
the nitrogen and sulfur contents
50 FEBRUARY 2021 | HydrocarbonProcessing.com
Process Optimization
of the different hydrocarbon the stripped water, meaning that uid/liquid coalescing equipment should
changing over time as the refinery it is bound to a non-volatile or also be considered as additional means to
processes different crudes, or strong acid in the stripped water further clean the sour water prior to the
different unit operations are added and will not come out of solution stripper. The SWS needs to be designed to
to the refinery. Ideally, the SWS can regardless of the energy input handle variations in inlet feed composition
handle most or all these changes into the bottom of the column. and flowrates, as well as allow a margin for
without major modifications to the In this case, it is prudent to install fouling and foaming. The selection of tray
stripper. A conservative estimate a nozzle in the lower section of internals should consider the severity of
of tray efficiency will provide the column to allow for caustic the service and the presence (or absence)
more flexibility in the design to addition, if necessary, under some of good sour water cleanup steps prior to
account for the uncertainty of or all conditions. The strong base the stripper. The tray design should take
the feed composition. will displace the NH3 and allow into consideration fouling, efficiency and
As previously mentioned, numer- it to be more easily stripped from hydraulics, among other factors.
ous SMEs use an initial rule-of-thumb the column. By placing the
tray efficiency of 33%, or three actual nozzle in the lower section Part 2. Part 2 discusses the design of
trays in the SWS for every equilibrium of the tower, the caustic will not packed tower SWS and details operat-
stage in the process simulation. To fur- interfere with H2S stripping. ing problems that can occur in an SWS
ther refine the cost estimate or proceed • Although trays can be designed for system.
with detailed design, it may be prudent fouling service, some reduction
to build a mass-transfer rate model of in efficiency will likely be noticed NOTE
the SWS. This can be more easily done over time. Even with adequate This article was originally prepared for the
once the column internals have been se- solids removal and hydrocarbon Brimstone Sulfur Recovery Symposium. It has been
edited and separated into two parts for publication in
lected, since accurate information about phase removal, some fraction Hydrocarbon Processing.
the tray—such as weir height, active tray of these materials will enter the
area, etc.—are critical to building an ac- column periodically. Some slightly- LITERATURE CITED
curate mass-transfer rate model. Reliable water-soluble hydrocarbons may 1
B. Scott, “Processes for Treating Refinery Sour
estimates of the sour water composition enter the tower and precipitate Waters,” Brimstone Sulfur Recovery Symposium,
will also be necessary to help ensure the in the lower section of the SWS Vail, CO, 1995
2
Asquith and A. Moore, “Sour Water Processing—
appropriateness of the SWS design. as the water heats to near boiling. Balancing Needs,” Brimstone Sulfur Recovery
Another important factor in the design Other salts may be present in the Symposium, Vail, CO, 2000
of the column is tray hydraulics. The ac- water that precipitate in the higher 3
D. Stevens, A. Mosher and D. Ogg, “Fundamentals
of Sour Water Stripping,” Brimstone Sulfur Recovery
tual hydraulics on the tray is dependent temperature areas of the SWS. Symposium, Vail, CO, 2007
on the tray device such as fixed-valve Adequate access to the column 4
J. Stavros and A. Keller, “Reducing Nitrates in
trays. The number of valves and size of for quick maintenance and some Refinery Wastewater,” Brimstone Sulfur Recovery
the opening is important to maintain liq- additional design margin may Symposium, Vail, CO, 2013
5
N. Hatcher, R. Alvis and R. Weiland, “Sour Water
uid on the tray and get proper contacting be prudent to address fouling Stripper Performance in the Prescence of Heat Stable
of the vapor and liquid; thus, the proper concerns. FIG. 4. shows an example Salts,” Brimstone Sulfur Recovery Symposium, Vail,
operating range for the design becomes of fouling that can occur in SWS CO, 2012
6
A. Keller, “Fundamentals of Sour Water Strippers,”
important. If the trays are overdesigned, service over a 5-mos period of Brimstone Sulfur Recovery Symposium, Vail, CO,
then the tray may weep or dump liquid, operation, which corresponded to 2015
resulting in poor operation. The design the 10%–15% reduction in vapor 7
P. Le Grange, “Operational Challenges in Sour
must account for the low end, as well as flow area noted by the authors.12,13 Water Stripping,” Digital Refining, online: www.
digitalrefining.com/article/1002335, July 2009
the high end of operations. One reason 8
D. Stevens and A. Mosher, “Fundamentals of Sour
24-in. tray spacing is often used is to give Takeaways. Stripping sour water is a de- Water Stripping,” Brimstone Sulfur Recovery
more capacity, especially when fouling or manding process in a refinery or gas treat- Symposium, Vail, CO, 2008
foaming is expected. ing facility. The sour water will contain a
9
McIntush, K., J. Farone, B. Piggot and C. Beitler,
“Specifying internals in sour water strippers,”
Even when the designer is confident multitude of contaminants in addition to Brimstone Sulfur Recovery Symposium (Virtual),
in the design of the column, some ad- the NH3and H2S stripped out of the water 2020
ditional precautions are recommended. in the process. These contaminants make 10
American Fuel & Petrochemical Manufacturers,
“AFPM Process Safety Bulletin—Hazards of Purged
These include: reliable operation of the SWS a challenge, Tanks—Formation of Pyrophoric Iron Sulfide in
• NH3 will be the more difficult but one that can be realized with appropri- Low Oxygen Environments,” No. 14-01, 2014
component to remove in most ate design of the SWS and the equipment 11
B. Spooner, “Reduce Hydrocarbons and Solids
sour water streams. NH3 has a high that surrounds it. Proper sizing and level Contamination in Sour Water Strippers,” VWR
International, 2013
affinity for water and will almost control of the three-phase separators in 12
R. Hauser and R. T. Kirkey, “Refinery Tests
always strip out of the sour water the sour water system are critical to re- Demonstrate Fixed Valve Trays Improve
after the H2S is almost completely moving contaminants—such as hydro- Performance in Sour Water Stripper,” New Orleans,
LA, AIChE, 2003
removed. It is possible to reach a carbons—that can severely impact SWS 13
F. Bela, Comprimo, “SWS Fixed Valve Trays,” 2003
stripped water condition where performance by causing fouling and foam- 14
N. Hatcher and R. Weiland, “Reliable Design of
the remaining NH3 is fixed in ing in the column. Solids filtration and liq- Sour Water Strippers,” PTQ , September 2012
Access to the handbook is included with all Full Access Subscription Plans
at HydrocarbonProcessing.com/Subscribe
zation reactor. The block flow diagram of to the RON of the total isomerate. There- solvent primarily used in the
a typical isomerization unit is shown in fore, the recycle stream should have req- extraction of edible oils. This calls
FIG. 2, as are the compositions of the light uisite amount of 2MP and 3MP to obtain for very high purity levels of hexane,
isomerate and the heavy isomerate stream. desired product specifications. However, followed by safe and careful storage.
The light isomerate is primarily com- this eventually leads to a buildup of the It also finds usage in the preparation
posed of four components: isopentane, third component of the recycle stream of rubber adhesives, can sealing
N-pentane, 2-methyl butane (2MB) and (N-hexane) in the recycle loop, affect- compounds, etc.
3-methyl butane (3MB). The focus on ing the throughput of the isomerization • Polymer-grade hexane (PGH):
improving the RON of this stream has reactor and creating a bottleneck when it PGH is a fast evaporating
made technology companies invest time comes to capacity augmentation. hydrocarbon solvent that consists
and money for further improvements This potential bottleneck led to the essentially of hexane isomers.
in processes and catalysts. The recycle novel idea of drawing N-hexane out of the A concentration of approximately
stream that is sent back to the reactor is recycle loop and using it to make a few 40% makes n-hexane the major
mainly a mixture of 2-methyl pentanes value-added products. N-hexane drawn component in this mixture.
(2MP), 3-methyl pentanes (3MP) and N- from the loop can be converted to mar- PGH is used as a polymerization
hexane. The lower the quantities of 2MP ketable products, such as food-grade/ medium and in the manufacture
and 3MP in the recycle stream, there is an polymer-grade/pharma-grade hexanes of catalysts.
increased tendency of the equilibrium to and isohexanes. Not only do these mar- • Isohexane: This compound,
shift in a direction to lower values of 2MB ketable products bring in additional rev- which can also be drawn from the
and 3MB in the product stream. These enue to the facility, but as the recycle to recycle stream, is a solvent used
are critical components that contribute the isomerization reactor decreases, more in industrial, professional and
feed can be pushed through the reactor. consumer applications, such as a
This is an added benefit for facilities, par- manufacturing process solvent,
Maximize Minimize ticularly where the isomerization unit is metal working and coatings. It is
2 Methyl butane 2 Methyl pentane bottlenecked. not sold directly to the public for
3 Methyl butane 3 Methyl pentane
FIG. 3 further describes the composi- general consumer uses; however,
tions of feed and product streams of the this product may be an ingredient
Maximize Minimize DIH column. in consumer and commercial
2 Methyl pentane Methyl cyclopentane product applications, such as
Feed
3 Methyl pentane ALTERNATE C6 PRODUCTS cleaning agents and coatings.
n-hexane
2 Methyl butane FROM THE ISOM UNIT • Special boiling point (SBP) spirit:
3 Methyl butane The recycle stream in an isomerization Depending on market demand,
2 Methyl pentane Maximize
3 Methyl pentane unit can be used to produce a variety of SBP spirit 55/115 can also be
n-hexane Methyl cyclopentane C6 products that are widely used in the produced from the recycle stream.
Methyl cyclopentane heavies
heavies
industry and are produced through alter- It is used in the rubber industry,
nate processing routes: particularly during the process of
FIG. 3. A closer look at a DIH column • Food-grade hexane (FGH): vulcanization in tire manufacturing
component split. Food-grade hexane is a colorless or in preparation of certain rubber
mixes, cements and adhesives. It is
also used as a thinner for varnish,
Maximize Minimize paint and printing inks formulation
where quick drying is required, and
2 Methyl butane 2 Methyl pentane as diluent for lacquer, enamels and
3 Methyl butane 3 Methyl pentane high-grade leather drops.
2 Methyl pentane TABLE 1 provides the typical specifica-
3 Methyl pentane tions of the products discussed here.
Revamps of conventional columns to flexible as the column is capable of oper- heat loads. The advantages of this process
DWCs can provide the following benefits: ating in two modes: the FGH mode, in are sufficient to prompt facilities to uti-
• Ideal alternative for revamp of which the column will produce a fourth lize it for beneficial production of FGH.
side-cut columns when high purity cut of FGH; and the DIH mode, in which FIG. 6 shows the process flow and the
is required from the three product the column operates in conventional components of the four cuts of the DIH
streams mode with recycle to the isomerization column post revamp. Benefits include:
• Lower footprint as equipment unit without FGH production. To pro- • Energy consumption is 30%
count is reduced by half duce the fourth cut, the alignment of the less than the conventional
• CAPEX and OPEX can be reduced wall inside the column is customized to column sequence.
by approximately 20%–50%. meet the desired product specifications, • With the drawing of FGH (i.e.,
Structurally, the exterior of a DWC quantities and also target to minimize the n-hexane from the recycle stream
looks like a conventional distillation
column, but inside a defining wall in the TABLE 1. Specifications for C6 products
column separates the tower into two sec-
Property Units FGH PGH SBP Isohexane
tions, creating different fractionation
zones. The zone in the column where the Color Saybolt Min. 30 Min. 25 Min. 30
feed is introduced works to effectively Density (at 20°C) kg/m 3
0.660–0.687 0.660–0.687 0.665–0.686
separate the heaviest and the lightest key. Moisture mg/kg 50 Max. 100 Max. 100
Because this wall removes the intrinsic Bromine index mg Br/100g Max. 50 Max. 10 Max. 5
mixing that takes place in the conven-
Refractive index 1.375–1.384 1.375–1.384 - 1.373
tional column by creating different sepa-
ration zones, these columns are thermo- Cu Strip Cor. <1 <1 <1
for 3 hr at 50°C
dynamically more efficient compared to
their counterparts—therefore providing Doctor test – Negative – Negative
benefits in terms of operating cost. Distillation range
Initial boiling point °C 65 64 50 59
DWCs in a DIH column for produc- Final boiling point °C 69 70 120 63
ing C6 cut. As demand for C6 products
Residue on mg/100 ml 5 5 1
surged, refiners foresaw an additional evaporation
source of revenue and hurried to gener-
ate C6 product. Components
The usual way of obtaining FGH N-pentane – – – Max. 1
from the recycle stream is by installing N-hexane wt% Min. 40 Min. 44 Max. 5
two new columns post the DIH column. Isohexane wt% 30–45 Min. 95
FIG. 4 shows the typical configuration of
Methyl cyclo pentane wt% Max. 20
producing FGH by the DIH route. In
this sequence, two new columns are in- Cyclo hexane wt% Max. 3
stalled downstream of the isomerization Benzene ppmwt Max. 500 Max. 3 Max. 100
recycle stream to produce FGH. Aromatics ppmwt Max. 10 Max. 3,500
An attractive alternative to this se- Lead as Pb mg/kg Max. 1 Max. 1
quence would be to revamp the existing Tota sulfur mg/kg Max. 5 Max. 2 Max. 1
DIH column using DWC technology to
produce four cuts.
For isomerization facilities, the re- Light
Light Light
isomerate isomerate
vamp of a DIH column to DWC can have isomerate
significant benefits. The revamped DIH
column produces light and heavy isom- Recycle Recycle
erate as top and bottom products, along Feed Feed Feed
with FGH and the recycle stream as the Hexane
Recycle product
other two cuts. FIG. 5 shows how a middle
DWC handles the overlap of the heavy
isomerate and the recycle stream, reduc-
ing the number of stages required for the Heavy Heavy Heavy
desired specifications compared to the isomerate isomerate isomerate
conventional column—the spare stages • DIH columns typically have 80–100 trays • The dividing wall reduces the stages • Spare stages become available
• Since the sidecut recycle stream is below requirement between feed and the for getting the 4th cut of hexane product
are available in obtaining the fourth cut. feed, there is considerable overlap of recycle stream for the same specs
This option of getting four cuts from heavy isomerate with the recycle stream • Spare stages available for the 4th cut
the DIH column is not only attractive in
FIG. 5. Use of DWC technology for a revamp of a DIH.
terms of lower CAPEX and OPEX, but is
Hydrocarbon Processing | FEBRUARY 2021 55
Process Optimization
distributor, and holes were found in the on this area, and all lines for the atmo- increases. This increase in pressure drop
downcomer where the top reflux dis- spheric distillation column were washed. revealed the possibility of flooding inside
charges. Maintenance carried out on the the column.
top reflux system was assumed to be the Problem description. The marked According to the findings of the first
reason for the oscillation problem. How- zones in FIG. 1 indicate the time inter- examination, with increased naphtha con-
ever, the problem continued after the vals during which the feed characteristic tent in the processed crude, the amount of
unit was restarted. changes and oscillation occurs. When overhead reflux increases to keep the over-
To determine the reason for the oscil- the top reflux amount rises above a cer- head temperature constant. This causes
lation problem, the unit was shut down tain value, the temperature oscillation be- fluid head to form in the area where flood-
again. This time, it was found that the gins. During this oscillation, the change ing is suspected. The liquid level formed
downcomer of the kerosene drawing area in pressure (ΔP) between the column exceeds the liquid carrying capacity of the
was plugged. Maintenance was performed top pressure and the flash zone pressure tray at some point, causing it to fall into
the lower trays, which makes them cool.
5,000 600 This situation creates the oscillation in
Overhead reflux amount, m3/d ∆P, overhead-flash zone pressure the product withdrawal temperatures.
500 A pressure survey was conducted to
4,000 ~ 4,000 m3/d
Overhead reflux amount, m3/d
∆P, g/cm2
300 gauges were installed in different zones
2,000 of the column, and the pressure data from
200 the points shown in FIG. 2 were recorded
1,000 twice per shift for one week. The recorded
100
pressure data included both conditions—
0 0 during oscillation and under operating
conditions without oscillation.
FIG. 3. The ΔP through the column and the overhead reflux amount. The change in crude oil content may
lead to an increase in the amount of over-
5,000 350 head reflux, depending on the ratio of
Overhead reflux amount, m3/d ∆P, overhead-kerosene stripper naphtha cut in the feed. FIG. 3 represents
300 the change in pressure drop value with
4,000 ~ 4000 m3/d respect to overhead reflux in the col-
Overhead reflux amount, m3/d
250
umn. It has been observed that when the
3,000 overhead reflux exceeds a certain level,
∆P, g/cm2
200
~ 200–220 g/cm2
the difference between the column over-
2,000
150 head pressure and the flash zone pres-
sure reaches approximately 370 g/cm2.
100 It indicates that liquid level accumulated
1,000 in the column gives rise to excessive pres-
50
sure drop through the column, since the
0 0 theoretical and anticipated pressure drop
is around 200 g/cm2–300 g/cm2.
FIG. 4. The ΔP between the overhead and kerosene stripper and the overhead reflux amount. FIG. 4 demonstrates the change in
pressure drop value obtained with the re-
5,000 300 corded pressure data from the transmitter
∆P, kerosene stripper/LAD stripper on the column top and from the pressure
∆P, LAD stripper/HAD stripper 250
4,000 Overhead reflux amount, m3/d ~ 4,000 m3/d
gauge installed on the kerosene stripper
200 with the amount of reflux. As the top re-
Overhead reflux amount, m3/d
During the inspection, a uniform con- to compare before-and-after values of critically high between the top and flash
tamination deposit was found on the the entire system. sections, but dramatically low between
the light diesel and heavy diesel
parts of the column. After mainte-
It is important to clean the column and nance, differences in pressure val-
ues, shown in FIG. 11, indicate that
to maintain steam-out during emergency the pressure was stabilized.
shutdown of the unit to guarantee operational
Takeaway. The blockage was de-
efficiency after maintenance is performed. tected at the downcomer of the ker-
osene drawing section of the atmo-
spheric distillation column. This
trays and on the reboiler exchanger shell As shown in FIG. 10, production draw blockage caused an increase in pressure
side of the stripper, as shown in FIG. 9. temperatures (especially kerosene draw at the top of column, an increase in tem-
During the shutdown of the unit, the temperatures) have stopped oscillating perature (depending on an instant rise in
atmospheric distillation column was and started to follow a straight line con- top load for lighter crude) and an increase
controlled from top to bottom, the side trolled by column dynamics and feed in top reflux due to a high top load. These
strippers were opened and checked for compositions, as desired. The atmospher- problems collectively resulted in temper-
deposits, and all of the drawing lines ic column is producing on-spec product. ature oscillation in the column. The unit
and internals were cleaned both manu- The overhead reflux value has also im- did not operate as expected at the maxi-
ally and with water. The CDU was then proved since the gas and liquid flows in mum unit charge.
successfully restarted and closely moni- the column have become uniform. When the contamination of the
tored. After enough data was collected As shown in FIG. 11, pressure values downcomer at the kerosene drawing sec-
from the running unit, the next step was during the oscillation in the column were tion was examined, the content of this
contamination was observed as 17% or-
ganic and 77% inorganic material. The
organic part was related to charge con-
tent, and the inorganic part resulted from
accumulation in the kerosene downcom-
er during turnaround blasting operations
in the column.
As learned from this study, it is impor-
tant to clean the column and to maintain
steam-out during emergency shutdown
of the unit to guarantee operational ef-
ficiency after maintenance is performed.
In addition, the control of the column
and its internals is crucial. During col-
umn control, the cleaning of the down-
FIG. 10. Before and after values of the overhead reflux and draw temperatures. comer sections and drawing lines must be
provided, and the drawing lines must be
6,000 600
washed until every line is clean.
Overhead reflux ∆, LAD/HAD
∆, top kerosene ∆P, top flash section GÖZDE ÖZTÜRK is a Process Engineer at Tüpraş’
∆, kerosene LAD 500 İzmit refinery, responsible for the crude and vacuum
5,000 distillation units. She previously worked as an
After maintenance Operations Engineer. Ms. Öztürk holds a BSc degree
400 in chemical engineering from Middle East Technical
Overhead reflux amount, m3/d
60 FEBRUARY 2021 | HydrocarbonProcessing.com
Maintenance
and Reliability
M. GREEN and R. RODRIGUEZ, CSNRI,
Riviera Beach, Florida
compatible composite repair system. These include: A newly required epoxy resin formulation was to be de-
• Mechanical, metallic clamp/patch/enclosure: Reasons signed for use as both a primer and saturant on repairs, and to
for using a compatible composite repair system include: meet recommended operational conditions. Additionally, the
° Cost—This option can be costly, depending on the chosen fibers had to be able to withstand the conditions. When
full scope of repair. a corrosive chemical comes into direct contact with a fiber, if
° Limitation on size—This option can have limitations the wrong type and grade of fiber are selected, then the chemi-
based on the size of pipe requiring repair, including cal can degrade the fiber and destroy the resin bond, result-
its overall weight. ing in a significant reduction in structural properties. Due to
° Lead times—This option can be associated with known reaction levels between sulfuric acid and carbon fibers,
long lead times, depending on the scope of repair. a glass fiber was chosen for the new system reinforcement.
• Cut and replace: A reason for using a compatible Among the primary goals that a development project for the
composite repair system includes: repair system should include are:
° Downtime—Time spent out of operation for this 1. Glass transition temperature of 130°C (266°F) or above
option can be extremely costly to the facility 2. Gel time longer than 1 hr
due to loss of production. 3. Lap shear strength greater than 580 psi (4 MPa)
• Existing composite repair systems: Reasons for using 4. Easy to mix epoxy parts A and B
a compatible composite repair system include: 5. Chemical resistance to sulfuric acid up to
° Compatibility—Most existing, commercially 98% concentration.
available composite repair systems are not compatible A stringent formulation and testing process was conducted
(or not proven to be compatible) with sulfuric acid to achieve the targeted properties through various epoxy for-
at concentrations above 40% or temperatures mulations. Various concentrations of base resins and advanced
above 24°C (75°F). hardeners were mixed and tested until the desired sulfuric acid
° Testing—Most existing, commercially available resistance, glass transition temperature (Tg ), viscosity and gel
composite repair systems do not have the physical times were achieved. Once the primary goal properties were
testing in place to validate claims of compatibility obtained, it was necessary to evaluate and adjust percentages of
in operating conditions. raw materials to meet all the specific performance goals. Vari-
ous fabric types, architectures and combinations were tested for
Development plan. Commercially available epoxy systems sulfuric acid resistance to find the final fiber-reinforced plastic
were reviewed for potential inclusion in the testing program; (FRP) makeup most compatible for sulfuric acid environments
however, despite claims of compatibility, there were other hur- with concentrations up to 98%. Results of the final formulation
dles that prevented them from being considered. The correct testing, along with some developmental discoveries of the sys-
selection and combination of the epoxy resin and the hardener tem’s curing protocols, are provided in the following section.
components determine the final characteristics and suitability Upon completion of the initial research and evaluation of a
of the system for a given environment, and, as such, a set of cri- suitable composite combination, full qualification testing to the
teria was developed for review of formulations. ASME PCC-2 Article 4.1 standard (and, similarly, to the ISO
24817 standard) was completed to fully qualify and character-
TABLE 1. Gel time test results ize the new system and its design properties and capabilities.
Sample no. Gel time (min)
Material evaluation. Firstly, a thorough research was con-
1 150
ducted to determine the appropriate epoxy formulation to be
2 137.4 used that would theoretically perform to the primary goals and
3 153
1.5
4 154
Average 148.6
1.0
1 1,150
2 1,140 0.0
3 950
4 1,132 -0.5 120 °C
5 1,023 134 °C(T)
142 °C
6 1,307 -1.0
7 1,266 0 50 100 150 200 250 300
Temperature, °C
8 1,295
Average 1,158 FIG. 2. Tg of epoxy, using DSC.
62 FEBRUARY 2021 | HydrocarbonProcessing.com
Maintenance and Reliability
expectations for the system. Upon completing this exercise and Tg determination. The Tg is the temperature range where a
formulating a sample batch of the determined formula, the next thermosetting polymer changes from a hard, rigid (or glass-
steps were performed to confirm that the required gel time, Tg like) state to a softer, more rubbery state. This is a critical value
and lap shear values were achieved with the formula. to know and understand for a composite repair system, as the
system must be able to maintain its mechanical properties—
Gel time determination. Gel time is the length of time that the consequently, upper temperature limitations should be known
two-part epoxy takes to gel after full mixing at a specified tem- to ensure proper function.
perature. This value can be useful to determine a working time of The Tg of the polymer being considered was measured
the system so that installers know the amount of time they need using differential scanning calorimetry (DSC) per ASTM
to install the product correctly. The test was performed with an E1356. FIG. 2 shows the Tg of the samples that were subjected
80-gram mass at 24°C (75°F), using a standard laboratory gel to a heat-cool-reheat cycle from 20˚C to 260˚C, with a ramp
timer for four different batches to achieve and confirm repeat- of 10˚C/min. The Tg obtained was 134°C (273°F), which
ability in the results. The average result for the measured gel time demonstrated that the goal for this property was achieved [i.e.,
was 149 min, demonstrating that the goal was achieved. TABLE 1 130°C (266°F)].
shows the results for the gel time for the different batches. In addition to these physical properties that were evaluated,
the “practical” requirement for ease of use was also considered
Lap shear strength determination. Adhesion testing, us- from the beginning. The requirement was for the components
ing a lap shear test, is used for a specific bond strength for the to be easy to mix, which meant that the viscosities should not
epoxy system to the metallic substrate. The testing was per- be such that they would be too high, creating a more difficult
formed according to ASTM D5868-08. Composite plates of mixture and causing difficulty in saturating the fiber. There was
the resin and fiber system were bonded to carbon steel plates not a specific target set, but an upper limit of 50,000 centipoise
with the epoxy of the system and, using a universal testing ma- (cps) was the expected maximum value. Upon mixing of the
chine, pulled in shear until failure of the adhesion occurred. epoxy, the final viscosity was measured to be approximately
The type of failure observed was cohesive. TABLE 2 shows the 25,000 cps, indicating a viscosity that could be successfully
results of lap shear strength for samples cured at 24°C (75°F). implemented within the system.
The average value of all eight samples tested was 1,158 psi Once each of these primary goals was achieved, the next step
(8 MPa), doubling the required goal. was to test for compatibility with the 98% sulfuric acid solution.
Technology
GasProcessi
ngNews.com
| SEPTEM
BER /OCTOBER
2019
intelligence for the
gas processing
industry
102
100
95
90
85
80
75
%T
70
65
60
55
50
45
1,447 1,400 1,200 1,000 912.51 800 600 495
cm–1
Name Cursor Description
Acid system 212°F – 19h_1 93.645 %T Sample 565 by Hanna Ford date Tuesday, April
Acid system 130°F – 9h_1 87.18 %T Sample 011 by Hanna Ford date Tuesday, January
FIG. 5. Composite panels after 1,000-hr of immersion and then cured at Acid system 78°F – 2h_1 69.735 %T Sample 564 by Hanna Ford date Friday, April
38°C (100°F) for 17 hr (left) and cured at 100°C (212°F) for 6 hr (right).
FIG. 6. FTIR spectroscopy of epoxy cured at different temperatures.
A view of the coupon sample before immersion and after 30 d
is shown in FIG. 3. be to post-cure the system at temperatures between 38°C
For the composite panel test, panels cured at various tem- (100°F) and 134°C (273°F), which is the established Tg of the
peratures and at different immersion times were also evaluated. epoxy system to speed the full curing of the system to a more
This allowed for a review of the effect of cure temperature on practical timeline. To characterize how this can be achieved
both short- and long-term immersion conditions. Results for the with target temperatures and time, a study of the cure schedule
different test panels and parameters are shown in TABLE 5. A view at various temperatures was completed.
of the composite panel before immersion, after 30 d of immer-
sion at a 38°C (100°F) cure temperature and after 30 d of im- Determining the cure schedule options. Two methods for
mersion at a 24°C (75°F) cure temperature is shown in FIG. 4. A evaluating levels of cure were used to determine the required
view of the composite panels after immersion for 1,000 hr, and curing time and temperature of the composite system: Shore D
then cured at 38°C (100°F) for 17 hr and cured for 6 hr at 100°C hardness studies, and Fourier-transform infrared (FTIR) spec-
(212°F), is shown in FIG. 5. troscopy tests to measure percent of reduction in epoxy peak.
The analysis of the change in appearance, weight, hardness The hardness of a composite system is a direct result of the resin
and tensile properties reflected “excellent” chemical resistance matrix type and how well it is cured. The more rigid the resin,
between 98% sulfuric acid and the composite system when al- the higher the level of hardness achieved—whereas, the more
lowed to cure for 7 d at both temperature levels. To obtain the flexible laminate will have a lower hardness level. Although hard-
best properties in the system, this cure schedule may be complet- ness increases with degree of cure, this method is not enough to
ed. It was noted in the 1,000-hr test that the panel cured at the fully analyze the cross-linking reaction. For example, it is pos-
lower temperature, while performing to a “good” rating, did not sible for an epoxy to appear solid, but to have not cross-linked to
perform as well as the panel cured at the higher temperature level, a degree that it will successfully resist the harsh chemical envi-
with regard to color retention (although mechanical properties ronment. Thus, the Shore D hardness study was complemented
were within the parameters of “excellent” rating), confirming that with the FTIR test to overlay the values and find the minimum
the level of cure of the epoxy polymer will still dictate the overall recommended cure schedule for the ECRS.
performance in long-term usage, and that the higher level of cure In the FTIR results, the presence of the epoxy group on the
achieved, the better the system will perform. If the polymer sys- infrared (IR) spectra is proven from the presence of a strong
tem is fully cured, which is based on time and temperature, then it band at 912 cm-1. Once the epoxy is fully cured, this peak should
will meet the requirements, but may take longer to do so. disappear, eliminating the area under the curve. It was demon-
However, due to requirements of how the system would be strated by the previous tests that, in these conditions, the system
used in a field environment, the recommended method would is chemically resistant to 98% sulfuric acid (TABLE 5). To ensure
Hydrocarbon Processing | FEBRUARY 2021 65
Maintenance and Reliability
90 80
Hardness over time
80 70 FTIR peak reduction over time
70 60
Area under epoxy peak
60 50
D-hardness
50 40
T = 75°F
40 T = 100°F
T = 130°F 30
30 20 T = 75°F
T = 100°F
20 10 T = 130°F
10 0
0 0 24 48 72 96 120 144 168
0 24 48 72 Time, hr
Time, hr
FIG. 8. Graph of FTIR peak reduction over time.
FIG. 7. Graph of hardness change over time.
ful at the previous time and temperature values, these were used
a fully cross-linked system (i.e., a fully cured system), the ulti- as the baseline to plot expected cure curves for the system. TABLE 6
mate Tg must be reached, which is achieved by curing the epoxy displays the correlation of time with Shore D hardness values and
at the measured Tg of 134°C (273°F). Note that successful usage percent reduction in FTIR peak values at three of the primarily
may not require full cross-linking, but only a specific percent- reviewed temperatures. FIGS. 7 and 8 show the change in hard-
age. While this study is not fully vetted in this article, it should ness and the FTIR peak, respectively, over time. As can be seen
be considered for the practical application capacity of a system. within the figures, the previous statement regarding hardness be-
This phenomenon can be seen in FIG. 6, where each line rep- ing only one piece of the puzzle for cure measurement is shown
resents the FTIR spectroscopy of a different cure temperature. to be true. Specifically, the samples cured at 24°C (75°F) reached
The top line represents a specimen cured at 100°C (212°F) for their full hardness in 2 d or less, while the associated FTIR peak
19 hr, where there is no epoxy peak at 912cm-1 because the resin measurements took up to 7 d to reach the minimum levels to be
is fully cured. The red line has a small peak at 912cm-1, which considered to have a resistance to the 98% sulfuric acid.
corresponds to a specimen partially cured at 55°C (130°F) for Based on this information, it can be determined that specific
9 hr. Finally, the blue line was cured at 24°C (75°F) for only curing protocols should be used for this system when it is going
2 hr; the depth of the epoxy peak means that there is not yet to be used with 98% sulfuric acid to ensure adequate resistance.
a crosslinking reaction between the epoxy and amine, indicat- The graphs can provide some ideas and indications of how to
ing that the sample is not fully cured. The area under the epoxy monitor and ensure that this cure level has been met when uti-
peak curve is what is considered for calculation purposes—the lized in the field for repairs.
smaller it is, the more cured the sample is until it disappears,
indicating a fully cured resin. Qualifying the final ECRS. Upon completing the validation
If this is not possible in the field of use conditions that this testing, and successfully characterizing the requirements and
ECRS will be subjected to, then a minimum cure must be known. meeting the primary goals set forth, the last phase is to complete
Since it was determined that chemical compatibility was success- the full qualification testing to the ASME PCC-2 Article 4.1 and
66 FEBRUARY 2021 | HydrocarbonProcessing.com
Maintenance and Reliability
ISO 24817 standards for ECRSs. While this article will not go izing and testing the composite materials across a spectrum of
into the details of each test, the summary of all the completed temperatures and other environmental conditions, users can be
qualification tests and achieved values are provided in TABLE 7. confident in the repair system’s ability to successfully function
as desired. The development of this new system is considered a
Discussion. This article provides a detailed insight into the breakthrough for its compatibility with sulfuric acid at 98% con-
process and steps followed for the development of a system that centration, as it is the first composite repair system fully tested
is resistant to sulfuric acid in concentrations of 98% at ASTM and proven to be resistant at such a level.
conditions of 24°C (75°F). The main objective of this project
was to create an ECRS that could be used as a valid and quali- NOTES
This article was originally presented at the 2018 NACE Middle East Corrosion
fied pipe repair option for pipe systems operating with 98% Conference
sulfuric acid, which has been achieved. Numerous discoveries
were made along the way—the most important of which was MATT GREEN is the Vice President of Technical Services at
CSNRI. He leads the company’s global technical department for
the effect of cure time and temperature on the resistance of the the composites business unit, which encompasses engineering,
system to the chemical solution. training and education. Since 2006, he has presented training
Upon full completion of the development and validation test- and education seminars around the world on non-metallic
ing of the system, the full qualification testing was completed on composite repair systems and their applications on pipelines and
piping networks. As a member of the Non-Metallic Composite
the system. In addition, once completed with the initial valida- Repair Subcommittee for ASME PCC-2 Articles 4.1–4.3, which writes and edits the
tion, subsequent chemical compatibility testing programs were ASME standards for composite repair systems related to the repair of pipelines and
started to widen the scope of chemical solutions that could be pipework, Mr. Green is active in shaping the industry and its future. He has
published numerous papers, and is a regular contributor of technical articles for
used compatibly with this system—regarding both chemical industry-related publications. He earned a BS degree in engineering physics from
compatibility and temperature ranges. Northeastern State University in Oklahoma.
Takeaway. ECRSs are being utilized as a routine repair option RUTH RODRIGUEZ is the Research and Development Project
Leader at CSNRI. She earned a Bch degree in chemical
for many process piping facilities, from refineries to fertilizer engineering and an MS degree in business administration from
plants to steel mills. Because of the harsh conditions found in Keiser University, Florida. She has deep experience in product
many of these facilities, advanced and thoroughly tested materi- development, bringing new products to the market and
developing epoxy resins for composite repair systems. During
als are required to be used successfully. Advancing technology the past decade, she has been focused not only on developing
requires commitment and persistence, as well as good coopera- and testing resins that offer the best mechanical properties and chemical resistance
tion between industry and the manufacturer. By fully character- to FRP systems, but also becoming an expert in GHS compliant SDS authoring.
FIG. 2. To ensure that the valves can operate reliably even in the
event of a fire, the actuators and actuator controls are equipped FIG. 3. Each metering skid has 12 actuators connected via a Modbus
with fireproof shields. Image courtesy of KROHNE. RTU loop with integral redundancy. Image courtesy of KROHNE.
68 FEBRUARY 2021 | HydrocarbonProcessing.com
Valves, Pumps and Turbomachinery
metering runs arranged in parallel, plus a master run for cali- The master station is particularly beneficial here, thanks to
bration. Actuators are needed for the flow control valves that its multiple redundancy options and its proven communica-
direct the oil or gas flow to the individual me-
tering runs and to the master run.
In one project, an automated metering Valve automation in metering skids for the oil
solution for a major oil and gas group in the
Middle East was implemented using a master and gas industry is challenging. Using a master
station that plays a key role in meeting the
a
station to control all of the actuators within
challenging requirements set by the custom-
er. In many cases, the supplier of the actuators the metering system has multiple benefits.
is dictated by the end user. However, in this
case, different actuators were introduced as a
cost-effective alternative in combination with the new, redun- tion via standardized Modbus protocols. For this project, a
dant master stationa. master station with hot standby system redundancy was sup-
Three metering skids were designed and manufactured, plied. Two subsystems, cost-efficiently located within a single
incorporating 36 actuators and one redundant master sta- housing, ensure that operation continues without interrup-
tiona for high-accuracy measurement of crude oil. Each of the tion, even if one subsystem fails (FIG. 3).
skids was equipped with three ultrasonic flowmeters. Within Communication to the distributed control system (DCS)
each skid, 12 actuators are used to control the flowrates and is also redundant, via Modbus TCP/IP (internet protocol
to switch between the different metering runs. They are con- suite). The actuators are connected to the master station via
nected to the master station in three separate loop topologies. Modbus RTU (remote terminal unit) in a loop topology. Re-
The master station provides central control and monitoring dundancy is also included at this level: if communication fails
of all 36 actuators across the three skids. To ensure that the at a specific position within the loop, then the master station
valves can be operated reliably even in the event of a fire, the considers both the resulting segments as individual lines and
actuators and actuator controls are equipped with fireproof all actuators remain accessible. Additionally, communication
shields (FIG. 2). via Modbus is extremely fast and efficient, resulting in short
cycle and reaction times.
High availability. Flow measurements from the metering The metering skids, including the valves and actuators,
skids are the basis for calculating the exact volume of oil or were fully assembled and tested at the manufacturing plant in
gas supplied and billed to the end customer. Any failure of the the Netherlands (FIG. 4). For transport to the end user, only
measurement system quickly results in high economic losses the external connections were separated, thereby minimizing
for the operator. High availability, therefore, was the custom- the steps needed to install and commission the pre-assembled
er’s prime requirement for the metering skids; this extended modules at the customer’s site.
to control and communication, as well as the reliability of the Using a master station as a central control hub for all of the
actuators themselves. actuators considerably reduces the external connections need-
FIG. 4. The metering skids were fully built and tested at the manufacturer’s plant in the Netherlands, including the valves, the actuators
and the master station. Image courtesy of KROHNE.
ed at the actuator level. Only two fieldbus cables are needed and the communication parameters set (FIG. 5). The actuator
per skid to connect the actuators’ Modbus loop to the master networks for each of the three metering skids also could be
station. At the customer site, only the master station then needs preconfigured using the master station. Commissioning was
to be connected to the DCS. accomplished with the master station’s automatic loop config-
uration feature, which facilitates actuator address assignment
Benefits during commissioning. Commissioning of the within the network.
actuators took place in the skids manufacturer’s factory us-
ing the master station, without requiring a connection to a Central diagnostic hub. During normal plant operation, the
DCS. The master station’s large, integral, multi-touch screen use of a master station reduces host communication to a mini-
allowed the actuators to be intuitively controlled and tested, mum, thereby reducing the workload on the DCS. The master
station distributes the individual operation commands to each
actuator, receives status updates from all of the actuators at
cyclic intervals, and transmits to the DCS only the concen-
trated data required for regular plant operation. In addition,
the master station offers a multitude of diagnostic functions.
Status and availability are clearly visible at all levels, from the
overall system down to each individual actuator. This speeds
up fault localization and repair (FIG. 6).
Additional customer requirements that evolved during the
course of the project were quickly accommodated. For exam-
ple, a new software feature to enable additional configurable
feedback signals was implemented. This feature is used to
monitor the external 24-V power supply to the actuators. The
master station also incorporates integral Ethernet interfaces,
FIG. 5. The master station operation is intuitive and convenient
opening new possibilities for enhanced data exchange, includ-
with either the integral, multi-touch screen or an external web ing Industrial Internet of Things (IIoT) applications.
browser. Image courtesy of AUMA.
Takeaway. The master station has provided added value for
metering skids due to easy commissioning, centralized con-
trol of all actuators, and a multitude of diagnostic functions.
Efficient, high-speed Modbus communication and integral
redundancy at all levels ensure reliable plant operation and
optimized processes.
NOTES
a
AUMA’s SIMA2 master station, used for metering skids built by KROHNE
Oil & Gas
SASCHA LOTH is SIMA2 Product Manager at AUMA Riester GmbH & Co. KG.
He has 20 yr of experience in designing and implementing automation solutions.
Mr. Loth holds an MSc degree in electrical engineering.
Listen to
Hydrocarbon Processing’s
podcast for the latest
technologies in the
downstream industry.
70 FEBRUARY 2021 | HydrocarbonProcessing.com
Supply Chain
K. R. IYER, M/s Indorama India Private Ltd.,
Haldia, India
information was then fed into a new ma- tinuous process industry professionals and updating the same in the aggregator
terial cataloging software to identify the and decision makers—even in this age of software was automated. Initially, some
commonality of spares among the com- computerization and e-commerce—that companies shared only the item master
details of surplus and non-
moving items while with-
Inventory sharing among manufacturing industries and group holding the sharing of SKUs
that were in regular use (gen-
companies can prevent costly outages, dispose of surplus and eral category). Also, many
obsolete inventory, and clean up the inventory by identifying companies did not share the
unit rate information. How-
and eliminating duplicate SKUs within the warehouse. ever, seeing the versatility of
inventory sharing in search-
ing the required inventory,
panies. Drawing up the common speci- common material cataloging is needed identifying the duplicates, generating in-
fication for the new codification system to identify the interchangeability among ventory reports such as inventory value,
and codifying the existing inventory took various units or factories. etc., the companies later began sharing
several months—the enormous time, It can be assumed that no two compa- the entire material master, including pric-
money and resources required to recodify nies have the same ERP system. It can also ing information. Based on the feedback
all stock-keeping units (SKUs) in a uni- be accepted that some companies may be and the frequency of lending and borrow-
fied way was a big stumbling block. It also properly maintaining the item master with ing among companies of costly spares and
required the companies to switch to the complete specification, make, model, size, generic items, the project was a success.
new unified system after going live. This material of construction, manufacturer
caused resistance from line managers as part number, etc., of the component or How to get started? Once the merits
everyone had to shift from existing ma- equipment where the component is fitted, of inventory sharing have been proven,
terial code to new code, which entailed while other companies will not have com- the unit head of the most influential
updating the item code at several places plete specification for the items. A differ- large company will normally take the
(bill of materials, purchase orders, mate- ence exists in the number of SKUs held by lead during interactions with counter-
rial tagging in warehouses, etc.) companies of various sizes, as well. parts in other companies. Once buy-in
Moreover, the sizes of the participat- The basic requirement of inventory is achieved, a point person (usually a
ing companies varied greatly: the num- sharing is that companies can continue maintenance or planning manager) is
ber of SKUs ranged from 15,000 to more to operate their day-to-day business using nominated from each company and the
than 100,000 in large companies. The their existing ERP and style of codifica- team leader is selected from among that
idea that the project’s cost should be tion, yet be able to quickly (in case of a group. The team should meet weekly to
shared equally was vehemently opposed stock out) access the inventory of a neigh- carry out the following steps:
by smaller companies, which began ques- boring company, find the required exact 1. Hold a kick-off meeting and share
tioning their participation. After almost a match or substitute, and borrow or buy it. contact details.
year of weekly meetings, the project was In a second attempt, it was decided 2. Collect data on the number of
shelved due to failure in finding a work- to pursue an aggregator software that SKUs and ERP systems of all
able solution. would collect the item master dump (in participating companies.
In another case, a large company with the prescribed format) daily from the 3. Review the method of
multiple units in its home country and ERP systems of participating companies, maintaining the item master and
overseas had a common ERP system and to update the same in a common ag- completeness of basic data.
across the group. The company wanted gregator software with a superior search 4. Decide which fields are to
all units to participate in inventory shar- facility. Proper data mapping was done be shared—the minimum
ing. Because all units shared the ERP for individual ERP systems in the aggre- information required are item
system, it was assumed that the only hin- gator software so that respective fields code, item short and long
drance to identifying the interchangeable were populated from flat files received description, manufacturer,
parts across the units was different codi- from various ERP systems as a part of the model number, manufacturer
fication systems. The task of implement- implementation scope. part number, unit of measure,
ing uniform material codes across all sites Not only were the required efforts re- stock on hand, surplus or
required significant effort. The individual duced, but the costs were also much lower disposable stock, size, rating,
units worked for several months to assign than the earlier model. The companies material of construction, unit
new codes and feed them into the ERP. agreed to participate and equally share rate, date of last purchase/
However, the result was unsuccessful, and the initial and yearly maintenance costs, receipt, item location, etc.
no inventory sharing took place. and the project was successfully commis- 5. Decide on optional fields to be
sioned in less than 4 mos after floating the shared, such as unit price, supplier
New approach revives project. These request for quotation (RFQ). name, purchase order price, etc.
examples of failures highlight the com- The process of extracting the item 6. Prepare an RFQ for inventory
mon misconception among many con- master dump from various ERP systems sharing aggregator software
72 FEBRUARY 2021 | HydrocarbonProcessing.com
Supply Chain
solution providers, including 10. Go live and allow team leads to panies can also identify multiple codes
a provision to view the test train other users of the ERP. and merge the duplicate stocks. In this
certificates, photographs of the The exercise can be completed in 3 example, Q-chem would send a request
component, etc. Send requests mos–4 mos and the payback can be less to Company A to sell the non-moving
through the software to all than 1 yr. The author has developed an SKU to them after verifying that the spare
team leads. There should be application for inventory sharing (FIG. 1) part is properly preserved by Company A
various ways to narrow down that can—in addition to the usual search by viewing the photo of the component.
the search (e.g., by make, model, criteria available in the aggregator soft- Company A will readily sell the item as it
size, rating, end connections, ware—locate the item even if it is mis- is non-moving with them. This is a win-
manufacturer’s part number, spelled. It can look for alternate dimen- win situation for both organizations.
supplier part number) and to sional standards (in. vs. mm, hp vs. kW,
search by synonyms, and accept etc.), display photographs of the item LITERATURE CITED
wild cards and keywords. Also, it selected, view detailed specifications and
1
Iyer, K. R., “Reduce inventory by eliminating mul-
tiple parts codes,” Hydrocarbon Processing, June 2004.
should be able to look for alternate test certificates, and more. 2
Iyer K. R., “Shape up your system for stocking spare
dimensions (e.g., if inches are As an example, suppose company parts,” Chemical Engineering, June 2004.
specified, it should also locate “Q-chem” is in urgent need of a bearing 3
Iyer K. R., “The secret of getting your inventory under
control,” Chemical Engineering, September 2005.
equivalent sizes, such as mm). shoe for its compressor and searches for
7. Arrange for a technical the part by entering the part number in K. R. IYER is Head of Manufacturing
presentation by the vendors that the software, as illustrated in FIG. 1. The Excellence for Indorama India Pvt.
have submitted their quotations. search has produced three SKUs, one held Ltd. and has more than 36 yr of
8. Negotiate and place the order to by Q-chem and two held by Company A. experience in the areas of projects,
reliability and maintenance, and
the selected vendor and decide There are duplicate SKUs in Company A: inventory control of large refineries
on the cost-sharing formula for one SKU is declared as non-moving and and fertilizer and petrochemical
software and annual licensing fees. stock is available for disposal; whereas the plants. He has authored numerous technical articles
9. Customize and upload the item second SKU of Company A shows a “gen- for various national and International journals. Mr. Iyer
holds a B.Tech degree in mechanical engineering
master dump and provide training eral,” meaning it is in use and required to and an MBA degree from South Gujarat University.
to team leads. be stocked and replenished. Thus, com- The author can be reached at kiyer@indorama.co.
.com
rocessing
arbonP
Hydroc
RY 2019 |
FE BR UA
HYDROC
ARBON
PROCESS
Inside intelligence
ING
IALS
MATERRING
FEBRUAR
CTU
for the downstream ANUFA
prove
Y 2019
can im
M Hy pe r- duplex
duplex
stainles
steel an
d
s steel
Ni -based
supera
lloy
ess cost
s
ery proc
industry
t refin
gers cu piping
MATER
e heat
exchan ent and
d- pl at cl ad equipm
IALS MA
Welde ion of
Fabricat
NU
FACTURIN
G
AND R n problems
io
e vibrat ter pum
p
Resolv e boos
in a crud in boile
r
actices t
Best pr en
VIE WPOIN
T
lue
st to va
from co nies
g safety d gas compa
Shiftin r oi l an
fo
reactor ignited, leading to five fatalities such as with the Occupational Safety and
Ope
sys
rati
ent
and major damage to the facility. While Health Administration’s (OSHA’s) pro-
on
gem
al d
the U.S. CSB investigation identified sev- cess safety management standard or other
ana
isci
Effective
m
eral other design and operating issues, hu- regulations. Occupational safety, industri-
plin
ety
programs
e
Saf
man errors like this must be anticipated al hygiene, environmental protection and
and appropriate safeguards provided to other programs may also be required to
help prevent serious injuries and other help ensure all hazards are appropriately
significant consequences. Focusing on identified and managed. Safety culture/leadership
OD—doing the right thing, every time— As shown in FIG. 1, effective safety pro-
is an essential component of maintaining grams consist of three interrelated foun- FIG. 1. Foundations of an effective safety
program.2
effective safety programs and achieving dations2 comprising:
Hydrocarbon Processing | FEBRUARY 2021 75
Environment and Safety
2. Safety systems—Comprehensive impact the effectiveness of an OD effort involved and passionate about
management systems provide a are listed below. Consideration should be safety and reflect the behaviors
framework to help ensure hazards given to implementing or strengthening they expect from their organization.
and associated risks are identified, safety and related programs in these areas A leader’s consistent behavior
evaluated and controlled.2,7 In many if they are deficient. helps build trust and engagement
cases, regulatory requirements Some precursors to effective OD pro- in the organization.
provide a good starting point grams include: • Employee engagement:
for defining the needed systems • Establishing safety as a core value Employees understand and value the
and requirements; however, in vs. potential conflicting priorities, importance of safe work activities
all cases, proper evaluation of such as cost or productivity and contribute to organizational
potential risks is required to help • Committing to consistent and programs and activities.
ensure they are controlled and visible leadership of safety • Procedural principles: Correct
managed appropriately. This • Promoting a sense of vulnerability ways of performing work are
may necessitate going beyond to support safety awareness and defined and completed as planned,
the minimum essential practice engagement and avoid complacency following documented and
defined by regulations, or may • Maintaining open communications authorized systems and procedures.
lead to complying with regulatory • Documenting operating procedures • Housekeeping and workplace
requirements when not required. and safe work practices standards: Standards are
Regulations [e.g., OSHA, U.S. • Developing effective and timely established for maintaining safe
Environmental Protection Agency training practices equipment, tools and facilities.
(EPA)], consensus industry • Implementing fitness-for-duty Employees are proud of their
standards [e.g., National Fire programs to mitigate possible worker work environment and
Protection Association (NFPA), impairment due to stress, fatigue, consistently maintain high levels
American Petroleum Institute alcohol, prescribed or over-the- of housekeeping.
(API)], industry guidance [e.g., counter medications, or illegal drugs Organizational OD efforts are closely
Center for Chemical Process Safety • Evaluating and managing related to good safety culture and leader-
(CCPS), American Society for process risks ship practices. As shown in FIG. 1, effec-
Safety Professionals (ASSP)] and • Monitoring and reviewing tive safety management systems also help
extensive literature provide detailed key performance indicators enable workers at all levels of an organiza-
guidance for implementing • Providing management of change tion to do their work correctly and safely,
appropriate safety management (MOC) programs to identify every time.
systems. and manage new hazards that The characteristics describing person-
3. Operational discipline— may be introduced al OD 2,11, as shown in FIG. 2, are intended
OD, as defined previously, relates to • Implementing equipment to support the day-to-day focus on OD to
how well safety and other systems inspection, testing and preventive help ensure all employees:
are followed.2,8,9,10 OD is influenced maintenance programs to help • Know how to perform their work
by many factors, including safety ensure tools and equipment are correctly and safely (knowledge)
culture and leadership, quality reliable and safe to use. • Commit and plan to perform the
of safety systems, and other work the correct way without
human behavior and human OD program characteristics. The deviations or shortcuts, based on
factors programs. characteristics and importance of effec- training (commitment)
Some precursors related to safety cul- tive OD programs have been discussed • Anticipate and are prepared
ture, leadership and safety systems that elsewhere2,11,12,13 and have recently been for what could go wrong or look
reviewed in the context of achieving ex- for and recognize what may be
cellent process safety performance,3,4 different in their current work
reducing loss of containment incidents14 environment and respond
and as a leading indicator of plant perfor- accordingly, based on training
Knowledge Commitment mance.15 OD programs comprise both and experience (awareness).
organizational and personal OD efforts. The goal is to have knowledgeable, pre-
The characteristics describing organiza- pared, experienced workers at all levels of
tional OD2,11 are intended to help com- the organization who account for the exist-
pany and/or facility leadership develop ing work environment rather than have an
effective OD programs, based on: unquestioning focus on strict adherence
Awareness • Leadership focus: Leaders to procedure when circumstances vary or
emphasize and provide a positive change. This requires developing appro-
work environment, managing priate operating procedures and effective
processes and resources for on-the-job training for required work ac-
effective programs and employee tivities, including recognition and trouble-
FIG. 2. Characteristics of personal OD.2,11
engagement. Leaders are personally shooting of possible deviations and the
76FEBRUARY 2021 | HydrocarbonProcessing.com
Environment and Safety
including procedures and training, It is necessary to start where you are, LITERATURE CITED
process safety information, alarms recognize where you are going and de- 1
Klein, J. A., “Operational discipline in the workplace,”
Process Safety Progress, Vol. 24, 2005.
and equipment functionality, and termine where you actually want to go 2
Klein, J. A. and B. K. Vaughen, Process safety: Key con-
instrument testing and calibration. in terms of improving OD. The FLAME cepts and practical applications, CRC Press, 2017.
• A lack of supervisory oversight model2 describes best practices for lead- 3
Klein, J. A., “Tune up process safety performance,”
and technically trained personnel ership efforts in getting started on im- Chemical Processing, October 2019.
4
Klein, J. A., “Sustaining effective process safety pro-
during the startup, an especially proving OD: grams in CPI facilities,” Chemical Engineering, February
hazardous period, was an omission • Focus: Develop a plan to provide 2020.
contrary to refinery guidelines. and enable appropriate focus on 5
U. S. Chemical Safety Board, “Vinyl chloride monomer
explosion,” Report No. 2004-10-I-IL, 2007.
No experienced supervisor or OD and effectively communicate 6
Center for Chemical Process Safety (CCPS), A practi-
technical expert was assigned to the plan and associated goals to cal approach to hazard identification for operations and
the startup after the day supervisor develop awareness and engage maintenance workers, Wiley-AIChE, June 2010.
left, although safety procedures site personnel, who are supported
7
Center for Chemical Process Safety (CCPS), Guidelines
for risk-based process safety, Wiley-AIChE, March 2007.
required such oversight. daily through leadership attention 8
Center for Chemical Process Safety (CCPS), Conduct
• An effective incident investigation and priorities. of operations and operational discipline, Wiley-AIChE,
management system to capture • Leadership: Leadership acts as 2011.
9
Klein, J. A., W. M. Bradshaw, L. N. Vanden Heuval,
appropriate lessons learned and visible role models, committed D. K. Lorenzo and G. Keeports, “Implementing an
implement needed changes had not to continuous improvement and effective conduct of operations and operational disci-
been employed. excellent OD performance through pline program,” Journal of Loss Prevention in the Process
• The mechanical integrity program the use of effective, consistent Industries, Vol. 24, 2011.
10
Vaughen, B. K., J. A. Klein and J. C. Champion, “Our
did not ensure that deficiencies leadership practices. process safety journey continues: Operational disci-
were identified and repaired prior • Accountability: Organizational, pline today,” Process Safety Progress, Vol. 37, 2018.
to failure, resulting in the “run to team and individual goals 11
Klein, J. A. and B. K. Vaughen, “A revised model for
failure” of process equipment. include a focus on OD with clear operational discipline,” Process Safety Progress, Vol. 27,
2008.
An effective OD program emphasizes expectations on performance, 12
Klein, J. A. and E. M. Francisco, “Focus on personal
completing all tasks correctly and safely, feedback and recognition, as operational discipline to get work done right,” Process
every time, regardless of the role of indi- appropriate. Safety Progress, Vol. 31, 2012.
13
Klein, J. A. and B. K. Vaughen, “Implement an opera-
viduals in the organization—it is impor- • Measurement: Metrics, audits tional discipline program to improve plant process
tant to recognize that OD is for everyone, and other tools are defined to safety,” Chemical Engineering Progress (CEP), Vol. 107,
not just operators. Kletz,18 for example, periodically assess site activities, 2011.
observed that: performance and progress
14
Klein, J. A. and S. Dean, “Develop a loss-of-con-
tainment reduction program,” Chemical Engineering
toward goals. Progress (CEP), Vol. 116, 2020.
Every accident is due to human error: • Engagement: Leaders provide 15
Bitar, F. K., D. Chadwick-Jones, M. Lawrie, M. Nazaruk
someone, usually a manager, has to a work environment that provides and C. Boodhai, “Empirical validation of operating
discipline as a leading indicator of safety outputs and
decide what to do; someone, usually engagement and support to plant performance,” Safety Science, Vol. 104, 2018.
a designer, has to decide how to do site personnel, based on good 16
U. S. Chemical Safety Board, “Refinery explosion and
it; someone, usually an operator, communication processes, fire,” Report No. 2005-04-I-TX, 2007.
has to do it. All of them can make employee input and involvement, 17
Baker, J. A., et al., “The report of BP U.S. refineries
independent safety review panel,” 2007.
errors, but the operator is at the end and interdependent behaviors. 18
Kletz, T., An engineer’s view of human error, 3rd Ed.,
of the chain and often gets all the Site personnel know they are CRC Press, 2001.
blame. We should consider the people important to success and their 19
Collins, J., Good to great: Why some companies make the
who have opportunities to prevent contributions are valued. leap and others don’t, HarperBusiness, 2001.
20
Collins, J., Turning the flywheel, HarperCollins, 2019.
accidents by changing objectives and Although OD improvement is ulti- 21
Clear, J., Atomic Habits, Avery, 2018.
methods, as well as those who actually mately the responsibility of everyone in an
carry out operations.18 organization, it can be helpful to assign an JAMES A. KLEIN, CCPSC, CPSA is a
OD champion and team to provide addi- Senior Consultant at ABSG
Consulting Inc. with more than 40
It is also important to recognize that tional focus and accountability for imple- yr of experience in process safety,
OD improvement efforts must relate to menting and sustaining effective OD pro- engineering and research. He
specific local site or organizational issues, grams. An important goal is to create an previously served as Process Safety
based on differing safety culture, lead- OD flywheel19,20 or habit 21, where initially Management Co-lead for North
America operations at DuPont. Mr. Klein is the
ership, work activities, hazards and/or small changes lead to larger improvements co-author of “Process safety: Key concepts and
geographic locations. These factors likely and momentum over time to achieve OD practical approaches,” published by CRC Press, and
vary from one site or area to another, es- goals and improve performance. has developed more than 50 publications, conference
presentations and university talks. He holds a BS
pecially in larger companies or facilities. degree in chemical engineering from the
Common issues may be identified, but lo- Part 2, which will appear in the March Massachusetts Institute of Technology (MIT), an MS
cal differences may lead to different prior- issue, will discuss raising OD awareness, degree in chemical engineering from Drexel
University, and an MS degree in management of
ities for improvement. Frequent observa- evaluating OD performance, pursuing technology from the University of Minnesota. He is a
tion of specific work activities can assist in improvement opportunities and sustain- Fellow of both the American Institute of Chemical
identifying potential OD improvements. ing OD programs. Engineers and the Center for Chemical Process Safety.
78 FEBRUARY 2021 | HydrocarbonProcessing.com
MIKE RHODES, MANAGING EDITOR
Mike.Rhodes@HydrocarbonProcessing.com
Innovations
Automating valve users and can help reduce service calls, By combining IT and OT data with ET
actuators maintenance and administration costs. data, SafetyInsight enables valuable engi-
The system allows users to monitor neering data (such as HAZOP and LOPA
Cowan Dynamics has made advance- Kidde Fire Systems addressable fire- reports) to be digitalized, and readily ac-
ments to its “C-PAC Module” Pneumatic suppression control units via computer, cessible by operation and maintenance
Manifold and launched the C-PAC store. tablet or smartphone. Using secure cellu- teams (FIG. 2) in simplified, intuitive and
The C-PAC (Cowan’s Pneumatic Actua- lar connectivity, control unit status along easy-to-understand visual formats.
tor Control) is an out-of-the-box solution with the status of all associated detection Ongoing risk assessments can then be
and can save up to 90% in assembly time devices and supervised suppression sys- based on digitalized safety data. The ad-
and easily automate valve actuators with- tems are at the user’s fingertips, providing dition of the IT/OT data provides near
out using cumbersome and costly piping information-based decision-making. real-time updates, enabling the accumula-
to connect various automation compo- Benefits to distribution channel part- tive impact of operation and maintenance
nents. The C-PAC Pneumatic Manifold ners and end users include: activities to be visualized on a dynamic
has the following capabilities for linear • Offers valuable insights with real- risk matrix, to further aid risk assessments
and 1/4 actuators: speed control, safety time event information and management of operational risk.
pressure relief and fail-safe capabilities. • Significantly reduces service times The suite delivers process safety dash-
The C-PAC Pneumatic Manifold’s with remote system analysis and boards to provide the right information,
preassembled manifold configuration maintenance planning
eliminates potential leak points and time- • Increases service levels and
consuming troubleshooting that normally customer satisfaction with 24/7
would occur with conventional manual in- remote access
stallations. Quality improvements such as • Reduces operational expenses
ASCO solenoid valves are now standard, and unbillable service calls by
replacing the option for generic brands. optimizing field service technicians’
The C-PAC is also available without time
pre-installed ASCO solenoid valves. Ac- • Keeps users informed of system
tuator technicians have the flexibility to status with automated notifications.
select any solenoid valve brand with the The IntelliSite system is compatible
appropriate area classification, ensuring with Kidde Fire Systems addressable con- FIG. 1. Kidde Fire Systems’ new IntelliSite™
complete compatibility. The online store trol units as well as legacy Chemetron® remote monitoring system for gaseous fire
gives valve automation centers and tech- and Fenwal® control units. The IntelliSite suppression systems.
nicians the ability to easily configure the mobile app is compatible with iOS™ 11
C-PAC module with more than 35 varia- and later, and Android™ 6 and later oper-
tions for their application. The C-PAC ating systems.
pneumatic manifold can be ordered in
three sizes: ¼-in. NPT, ½-in. NPT and Digitalizing process safety
1-in. NPT.
lifecycle management
Monitoring solution ABB has launched ABB Ability™
SafetyInsight™, a suite of digital software
for gaseous fire applications that supports companies
suppression systems across the energy and process sectors
Kidde Fire Systems has launched its throughout the entire lifecycle of pro-
new IntelliSite™ remote monitoring sys- cess safety management (PSM). Oper-
tem for gaseous fire suppression systems. ating as a central source of information,
The IntelliSite system allows users to the software digitalizes early engineering
monitor the status of a portfolio of fire technology (ET) data to create a process
control units in real time across multiple safety digital twin, giving context to the
locations. With 24/7 information access, vast amount of data generated through FIG. 2. ABB has launched ABB Ability™
the IntelliSite system (FIG. 1) enhances information technology (IT) and opera- SafetyInsight™, a suite of digital software
applications.
safety and service levels for system end- tional technology (OT) systems.
Hydrocarbon Processing | FEBRUARY 2021 79
Innovations
to the right person, at the right time, to heat trace controllers with the control communications are prohibited, the
make the right, informed decision, ac- room using wireless communications. network can be configured using a tradi-
cording to ABB. The solution presents alarms, history tional wired solution.
and operational data via a user-friendly,
Managing, connecting browser-based interface accessible from Handheld thermal
any network-connected computer or
and optimizing mobile tablet. Utilizing an adaptive, imaging cameras
heat trace systems self-healing wireless mesh network and FLIR Systems Inc. has made four new
Thermon Group Holdings Inc. has event-driven communications, users ex- additions to its Exx-Series of advanced
presented the Genesis™ Network, a new perience reliable, real-time status report- thermal imaging cameras (FIG. 4): the
solution providing site-wide operational ing and responsive control of their heat E96, E86, E76 and E54. Compared to
awareness and supervisory control of trace systems. predecessor Exx-Series cameras, the new
heat trace systems. With the Thermon The following new components work cameras offer enhanced thermal resolu-
Genesis Network (FIG. 3), users save together to form a Genesis Network: tion for more vibrant, easy-to-read im-
maintenance hours, increase up-time, Genesis Server, an on-premises or cloud ages and on-camera routing capability
and make upgrades and changes with deployable server; Genesis Gateway, a to improve field survey efficiency. The
greater flexibility. Users can easily moni- communications device between the serv- new Exx-Series cameras are designed to
tor, maintain and troubleshoot even the er and wireless mesh; and Genesis Bridge, help professionals detect the early signs
largest heat trace systems that may in- a communications device between the of building issues, identify hot spots,
clude more than 10,000 heat trace cir- mesh and heat trace controllers. troubleshoot electrical and mechanical
cuits. The Genesis Network connects Maintenance personnel need timely systems, and prevent problems before
and complete awareness of the trace heat- they cause damage that leads to expen-
ing system to prevent downtime and op- sive repairs.
timize operations. They also require in- The E96, with a 640 × 480 resolution
formation to troubleshoot issues quickly and eight-times digital zoom, is the most
and accurately. The Genesis Network advanced Exx-Series thermal camera
answers these challenges by delivering to date. It delivers improved measure-
alarm management, setting optimization, ment results over the greatest distance
historical analysis and operational re- to target, so professionals can safely di-
porting. The solution can save significant agnose electrical faults or locate hidden
time through the use of historical data to anomalies at very high temperatures up
optimize settings and accurately flag out- to 1,500°C (2,732°F), including in harsh
lier behaviors while avoiding nuisance industrial environments.
alarms, resulting in fewer maintenance For the first time, FLIR Inspection
hours. The Genesis Network also saves Route is now offered as a standard fea-
FIG. 3. The Thermon Genesis™ Network time by filtering and prioritizing alarms ture on every Exx-Series camera and is
provides site-wide operational awareness and and generating reports with the right in- complemented by the FLIR Thermal
supervisory control of heat trace systems. formation to make efficient decisions. Studio Pro software with Route Creator
The Genesis Network is architected plugin, sold separately as an annual sub-
as a true Industrial Internet of Things scription. The complete routing bundle
(IIoT) solution. An intuitive browser- enables professionals to create and ex-
based user interface gives users quick ac- port custom inspection and pre-planned
cess for performing alarm management, routes, ideal for large or multi-location
report generation, data trending and per- electrical or mechanical projects.
formance optimization of heat trace cir- The E96, E86 and E76 include Ul-
cuits while on-the-go from any network- traMax® high-definition image enhance-
connected laptop or mobile tablet with ment technology and improved contrast
a browser. Adding or changing devices with one-touch level and spanning func-
is a simple drag-and-drop activity. Soft- tions to view greater image details. In ad-
ware updates to the network can also be dition, interchangeable AutoCal™ lenses
deployed to all devices via the browser offer complete coverage of near and
interface, making the roll out of new fea- distance targets, with the built-in laser
tures simple, increasing the value of the distance meter ensuring the crisp focus
solution over time. needed for accurate temperature mea-
Designed with security in mind, from surement.
data encryption and authentication to
FIG. 4. FLIR Systems Inc. has made four secure updates, operational leaders can An expanded version of
new additions to its Exx-Series of advanced deploy the Genesis Network with peace Innovations can be found online
thermal imaging cameras. of mind. In applications where wireless at HydrocarbonProcessing.com.
80 FEBRUARY 2021 | HydrocarbonProcessing.com
Catherine Watkins, Publisher ADVERTISER INDEX
Phone: +1 (713) 520-4421
Catherine.Watkins@HydrocarbonProcessing.com
www.HydrocarbonProcessing.com 2020 Catalyst Handbook ..........................................52
BRAZIL
Evan Sponagle
Phone: +55 (21) 2512-2741
Mobile: +55 (21) 99925-3398
Evan.Sponagle@GulfEnergyInfo.com
Events
82 FEBRUARY 2021 | HydrocarbonProcessing.com
CALL FOR ABSTRACTS IS NOW OPEN!
H2-TECHSOLUTIONS.COM #H2TECH
LO-CAT:
Liquid redox
system for
H 2 S removal.
Convert H 2 S into innocuous,
elemental sulfur, with up
to 100-percent turndown
and no hazardous waste.
That's innovation for the
bottom line.
TREATING TECHNOLOGY
Caustic treatment design and waste disposal for gas plants
Increase separation efficiency with enhanced slug catcher internals
TURBOMACHINERY
Improve turbine operation with a reliability analysis
BACK TO BASICS
Natural gas liquids extraction and separation
Special Supplement to
Comprehensive Energy Project Data to
Help You Grow Your Business … Again
HYDROGEN
Global Energy Infrastructure (GEI) is a new market intelligence tool for the international
energy industry. With the recent disruptions across all energy markets, it’s more important
than ever to get up-to-date information on projects, including project status.
Combining our well-established downstream project data set, used by the biggest names
in global oil and gas, with new data sets for hydrogen, renewables, global pipelines, and
LNG, GEI gives you the intelligence you need to grow as the energy industry recovers and
transitions.
Designed with ease-of-use in mind, GEI equips you and your team with access to the
latest data in an easy-to-access and easy-to-use format. All data and market intelligence
reside on the GEI website, updated 24/7/365 by our researchers in Houston, London, and
Mumbai.
For customers who need the data in spatialized formats, Energy Web Atlas
continues to provide a superior tool which maps global energy assets for
midstream, downstream, renewables, and hydrogen.
11
This year, the Gulf Energy Information
family expands with the introduction of
H2Tech, the first technical publication de-
A� BLUME, voted entirely to applications and trends
Editor-in-Chief for the hydrogen (H2 ) community world-
wide. H2Tech is geared toward engineers
and other professionals involved in H2 production technolo-
gies and applications progress on all spectrums—from gray
and brown to blue and green, and beyond.
The midstream sector is reducing its carbon footprint with
help from H2 blending in natural gas pipelines, the use of H2 in
turbomachinery and, in some cases, the repurposing of exist- SPECIAL FOCUS: TREATING TECHNOLOGY
ing natural gas infrastructure to distribute H2 . Forecasts show 11 Caustic treatment design, integration
that H2 could reach a 10% share in the world’s total primary and waste disposal for gas plants
energy demand by 2050. J. Trucko and J. Tertel
H2Tech was created to inform and unite the community of
technologists and business professionals working in the rapidly 15 Increase separation efficiency
growing H2 sector. The first quarterly issue of H2Tech will be with enhanced slug catcher internals
published in March 2021, and the website will host news, pod- S. Alharbi, T. Alghamdi and A. Alhazmi
casts, webcasts, whitepapers, engineering resources and much
more. We invite all readers to visit www.H2-Tech.com to stay 19 Technical and economic impacts of
updated on the H2 sector and to sign up for the H2Tech weekly piperazine content in MDEA solvents
newsletter. On 18–19 May, the H2Tech Solutions virtual tech- S. Williams, C. Ridge and D. Engel
nology conference will bring together engineers, technologists
and managers working to advance fuel, chemical and industrial TURBOMACHINERY
applications for H2 . Visit www.H2-TechSolutions.com to regis- 25 Improve gas turbine operation
ter and submit your abstract. Finally, please feel free to drop me a with a reliability analysis
line with feedback or editorial inquiries for H2Tech at Adrienne.
A. Hadroug, A. Hafaifa, M. Telha and A. Iratni
Blume@H2-Tech.com. I look forward to hearing from you! GP
Other Gulf Energy Information titles include: Hydrocarbon Processing ®, World Oil ®,
Petroleum Economist ©, Pipeline & Gas Journal and Underground Construction.
GAS PROCESSING NEWS
A. BLUME, Editor-in-Chief
U.S. FERC approves ADNOC, Total mark first unconventional gas from UAE
Alaska LNG The unconventional
import project gas was delivered from the
Ruwais Diyab unconventional
U.S. energy regulators gas concession, located
approved a plan by Marathon 200 km west of Abu Dhabi
Petroleum Corp.’s Trans-Foreland city, in November 2020.
Pipeline Co. LLC unit to convert The achievement marks a
the Kenai LNG export plant in significant milestone toward
Alaska into an import terminal. future full field development
Trans-Foreland has 2 yr to put and is an important step
the proposed project into service. toward ADNOC’s target of
The facility would import producing 1 Bsft3d of gas from
up to four tanker loads of LNG the concession before 2030,
per year and use its boil-off gas ultimately enabling gas self-
management system to deliver sufficiency for the UAE.
imported gas to the adjacent The unconventional gas is
Kenai refinery. The Kenai LNG delivered through a purpose-
export plant entered service in built gas pipeline and centralized early production facility in the Diyab field, which enables distribution
1969, but it has not exported through ADNOC’s gas network. The accelerated progress and strong collaboration between ADNOC and
LNG since 2015. Total enabled the companies to fast-track the exploration of the unconventional gas resources while tailoring
operations to the UAE’s shale play type.
Nikkiso Cryogenic The milestone builds on ADNOC’s continuous efforts to de-risk unconventional gas resources across
sells turbo line Abu Dhabi since 2016 and comes just over a year after the country announced the discovery of 160 Tsft3 of
unconventional gas recoverable resources.
to Air Liquide
Nikkiso Cryogenic Industries’ Sempra to move forward Azerbaijan
Clean Energy and Industrial Gases
Group, part of Nikkiso Co. Ltd. with Costa Azul LNG starts gas exports
of Japan, announced the sale
of its Turboexpander Business
Sempra Energy’s ECA to European
Liquefaction subsidiary will
Line to Air Liquide in January. build the Costa Azul LNG export market
The Turboexpander Business plant in Mexico, the only LNG Azerbaijan has started
Line designs, manufactures and export project in the world to commercial natural gas supplies
sells turboexpanders within the get a final investment decision to Europe via the Trans Adriatic
industrial gas and NGL industries. (FID) in 2020. The government Pipeline (TAP), beginning its push
Nikkiso’s Cryogenic Service awarded the export permit on into the lucrative energy market
unit will remain an authorized the condition that the project will help offset an oversupply of gas dominated by Russia. The project
service company and will in the area. Sempra had been waiting for the export permit all year, has the backing of the European
continue to provide aftermarket due to delays caused by the spread of the COVID-19 pandemic. Commission as part of efforts to
services, including repair and ECA Liquefaction is a JV between Sempra LNG and Sempra’s curb Europe’s dependence on
servicing of ACD designed and Mexican subsidiary, Infraestructura Energética Nova SAB de CV Russian energy. Russia controls
built turboexpander machines, (IEnova). The $2-B terminal is expected to produce first LNG in late 34% of Europe’s gas market
while Air Liquide will provide 2024. The plant, which will have a nameplate capacity to produce and plans to raise gas exports
service activities to its plants and about 3.25 metric MMtpy of LNG, already has 20-yr agreements to Europe, including Turkey,
its third-party plant customers. with units of Mitsui & Co. Ltd. and Total SE for the purchase of to 183 Bm3 in 2021 from the
The acquisition was effective 2.5 metric MMtpy from the project’s first phase. 171 Bm3–172 Bm3 expected
January 1, 2021. Costa Azul’s Pacific Coast location gives it an advantage over for 2020. Europe’s annual
competing U.S. Gulf Coast export plants because it is closer to growing consumption stands at around
Asian markets. U.S. terminals usually ship LNG to Asia through the
Double E seeks Panama Canal.
500 Bm3, used mostly in
power generation.
approval for TAP is a part of the $40-B
Texas-New Mexico O&G industry commits to new Southern Gas Corridor, stretching
3,500 km from Azerbaijan
gas pipeline methane emissions reporting to Europe and drawing from
Double E Pipeline LLC has In a move that will help tackle one of the largest and most solvable Azerbaijan’s giant Shah Deniz
asked U.S. energy regulator FERC contributors to the climate crisis, 62 major players in the oil and gas II field in the Caspian Sea.
for permission to start building industry that represent 30% of global oil and gas production have agreed Azerbaijan aims to supply
its natural gas pipeline project in to report methane emissions with a new, higher level of transparency. European gas markets with 10
Texas and New Mexico. Double The Oil and Gas Methane Partnership (OGMP) is a Climate and Clean Bm3y of gas, including 8 Bm3 to
E is owned by units of Summit Air Coalition initiative led by the UN Environment Program, the European Italy and a combined 2 Bm3 to
Midstream Partners LP (70%) and Commission and the Environmental Defense Fund. At the core of the Greece and Bulgaria. It already
ExxonMobil Corp. (30%). effort is a comprehensive, measurement-based methane-reporting supplies gas to Turkey.
Double E Pipeline LLC plans to framework (OGMP 2.0) that will make it easier for officials, investors Azeri gas is unlikely to
build a 135-mi (217-km) pipeline and the public to accurately track and compare performance across change Russia’s dominant
to transport 1.35 Bft3d of gas companies in ways that have not been possible to this point. position in Europe, but it still
from the Delaware Basin in the Crucially, the OGMP 2.0 includes not only a company’s own poses some threat to Russia’s
Permian Shale in New Mexico and operations, but also the many JVs responsible for their production. share of the EU gas market.
Texas to the Waha Hub in West The OGMP 2.0 framework applies to the full oil and gas value chain, Russia, meanwhile, is seeking to
Texas. Double E is one of several including midstream transportation and downstream processing and complete its subsea Nord
pipelines proposed to transport refining. To support the realization of global climate targets, OGMP 2.0 Stream 2 gas pipeline project
Permian gas and is expected to aims to deliver a 45% reduction in the industry’s methane emissions to Germany, which stalled a
enter service in 2021. by 2025, and a 60%–75% reduction by 2030. year ago due to U.S. sanctions.
4 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
OH YEAH,
WE HAVE A FILTER FOR THAT.
No doubt about it. Whatever process you’re running, we have a filtration and separation
solution that will save you time and money in the long run. And if your process design
has you stretching out into new territory, don’t worry. New territory is kind of our thing.
713.849.0849 | ftc-houston.com
GLOBAL PROJECT DATA
LEE NICHOLS, Editor-in-Chief/Associate Publisher, Hydrocarbon Processing
Hydrocarbon Processing’s Construction Boxscore Database the COVID-19 virus, new gas processing/LNG project an-
and Energy Web Atlas are tracking 330 active gas processing/ nouncements showed a drop in the last two months of 2020.
LNG projects in the hydrocarbon processing industry. Most However, with COVID-19 vaccines being distributed, 2021
of these projects—approximately 35%—are in the Asia-Pa- may be a rebound year in natural gas consumption, resulting
cific region, followed by the U.S. (21%). Due to additional in additional capital being spent to supply increased demand,
lockdowns and restrictions stemming from a resurgence in especially in Asia. GP
23
Canada 41
22
70 Eastern Europe,
Western Europe Russia, CIS
U.S. 25 117
23
Middle East
Africa
9
Latin America Asia-Pacific
Detailed and up-to-date information for active construction projects and operational assets
in the gas processing, LNG and pipeline industries across the globe | EnergyWebAtlas.com
6 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
SUBSCRIBE TO THE LATEST
TECHNICAL NEWS IN HYDROGEN!
BE PART OF THE
HYDROGEN REVOLUTION
H2Tech is a new initiative to better serve the hydrogen sector. Through
a quarterly technology journal, weekly e-newsletter, podcast and virtual
conference, H2Tech is devoted to applications and trends for the
hydrogen community worldwide!
H2TECH INCLUDES:
» Technology advances for » Safety and sustainability
hydrogen production and use » Market pricing trends
» Blue and green hydrogen projects » The future of hydrogen energy
» Hydrogen distribution and fueling » And much more!
infrastructure development
H2-TECH.COM #H2TECH
INDUSTRY METRICS
ADRIENNE BLUME, Editor-in-Chief
U.S. exports of LNG set a new record in December 2020, averaging U.S. gas production (Bft3d) and prices ($/Mcf)
9.8 Bft3d, according to the U.S. EIA. The December volume is more 120 7
than three times higher than the reduced export levels seen in
100 6
Production, Bft3d
as higher demand in key Asian markets on colder-than-normal winter 80
temperatures and unplanned outages at LNG export terminals in 4
60
Australia, Qatar, Malaysia, Norway, Nigeria and Trinidad and Tobago. 3
40 Monthly price (Henry Hub)
The reduction of LNG supplies led to upticks in natural gas and LNG 2
20 12-month price avg.
prices in Asia and Europe, attracting higher volumes of flexible LNG Production 1
supplies from the U.S. GP 0 0
MAM J J A S OND J F MAM J J A S OND J F MAM J J A S ON
2018 2019 2020
U.S. natural gas spot prices at Henry Hub Production equals U.S. marketed production, wet gas. Source: EIA.
15
40
$/MMBtu
NGL
LPG
10 30 Ethane/ethylene
Propane/propylene
5 20
10
0 Oct.- Nov.- Dec.- Jan.- Feb.- Mar.- April- May- June- July- Aug.- Sept.- Oct.-
Mar. April May June July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. April May June July Aug. Sept. Oct. Nov. Dec. Jan. 2019 2019 2019 2020 2020 2020 2020 2020 2020 2020 2020 2020 a2020
2019 2019 2019 2019 2019 2019 2019 2019 2019 2019 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2021 Source: U.S. EIA
pipeline-opportunities.com
8 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
REGIONAL FOCUS
Advancing LNG infrastructure in West Africa. Investment The previous approval of a concession by Tema LNG Ter-
in LNG import infrastructure is an emerging trend in West Af- minal Co. Ltd. (Tema LNG), a JV between Helios Investment
rica as leading natural gas producers such as Nigeria, Ghana, Partners and Ghana National Petroleum Co. (GNPC), for the
Côte d’Ivoire, Senegal, Equatorial Guinea and Benin grapple construction of an LNG import terminal was a major step in
with challenges that have constrained the consistency of their Ghana’s efforts to hedge available gas supplies for power gen-
gas production. eration. GNPC previously signed a 12-yr gas supply agreement
The region’s natural gas producers are still courting interna- with Rosneft Trading SA for the supply of 1.7 metric MMtpy of
tional oil companies with opportunities in their upstream spaces, gas to serve the Tema region.
but production has remained inadequate to support increased Under the supply agreement, GNPC will purchase LNG
power generation requirements. This has led to a shift in focus from Rosneft Trading for regasification and subsequent deliv-
to the construction of LNG regasification and storage facilities. ery to thermal independent power producers and industrial
For example, Nigeria, Côte d’Ivoire, Ghana and Senegal— customers. In accordance with the plan, Rosneft Trading has
the leading gas-to-power markets in West Africa—have plans signed an agreement with Tema LNG Terminal Co. for the de-
to generate a combined additional 7,759 MW of power from velopment of a 240-MMsft3d LNG import terminal that will
gas by 2030, of which 2,222 MW is projected to come online be linked to a floating storage unit (FSU) and a floating re-
by 2025. To realize this plan, the countries’ governments have gasification unit (FRU). The FRU arrived offshore Ghana in
entered into agreements with international gas suppliers to early January 2021.
supplement their inadequate domestic supply. They have also The Tema LNG terminal project was preceded by a similar
drawn up LNG storage projects to meet the seasonal demand deal between GNPC and Bermuda-based LNG shipping com-
expected from end users, such as electricity generation plants pany Golar for the supply of 170,000 m3 of gas from the Golar
and industrial users. Tundra FSRU to support Ghana’s LNG imports.
Ghana pushes ahead with Tema LNG. Ghana is targeting Côte d’Ivoire eyes FSRU. In Côte d’Ivoire, where gas is a criti-
an additional 1,718 MW of power generation from gas by 2030. cal part of the country’s energy mix, the government is working
In partnership with the private sector, the government has in- to fast-track modern LNG import infrastructure to regasify and
vested in new commercial LNG import infrastructure, namely store LNG for distribution to end users.
the Tema LNG import terminal (FIG. 1). At present, Ghana re- The country has an estimated 1 Tft3 of developed gas resourc-
lies on the 678-km West African Gas Pipeline for its gas supply; es, but this is inadequate to meet demand; Côte d’Ivoire has an
however, the pipeline has a historically erratic supply trend due estimated gas deficit of 0.35 Tft3y. Additional gas must be im-
to payment disputes, technical issues and upstream gas supply ported, regasified, stored and distributed to support the planned
issues with Nigeria. generation of 1,271 MW of gas-fired electricity by 2030.
Gas Processing & LNG | JANUARY/FEBRUARY 2021 9
REGIONAL FOCUS
Earlier, French oil major Total led a team of partners to sign tanks with a capacity of 14,000 m3 of storage capacity. The re-
an agreement for the development of a 3-MMtpy LNG regasifi- gasification plant, located at the Akonikien port, is the first LNG
cation terminal in Côte d’Ivoire. Other companies involved in storage and regasification plant in West Africa and falls under
the LNG project, in which Total is the operator with a 34% Equatorial Guinea’s government-led LNG2Africa initiative.
stake, include national firms PetroCI (11%) and CI Energies Meanwhile, an 84,100-m3 LNG storage tank is being de-
(5%), as well as SOCAR (26%), Shell (13%), Golar (6%) and veloped in Nigeria as part of the $6.5-B Nigeria LNG Ltd.
Endeavor Energy (5%). (NLNG) Train 7 project, which also includes a new liquefac-
Total will supply up to 0.5 metric MMtpy of LNG for the tion unit, a 36,000-m3 condensate tank and three gas turbine
regas terminal over a 15-yr period, making it possible for Côte generators on Bonny Island. The development of gas storage
d’Ivoire to integrate LNG supply and import infrastructure facilities is a boost for Nigeria’s gas-to-power scheme, partly
through an FSRU. A pipeline would connect the FSRU in the because of the uncertainty of gas supply from key producing ar-
Vridi, Abidjan area to existing and planned power plants in Abi- eas, such as the Niger Delta, where frequent vandalism starves
djan, as well as to regional markets connected to the Ivorian net- power plants of the needed gas. Nigeria selected a consortium
work. Total said the project will enable Côte d’Ivoire to become comprising Saipem, Chiyoda and Daewoo as the EPC contrac-
the first regional LNG import hub in West Africa and to meet tor for the Train 7 project.
both regional and domestic demand. One of West Africa’s smallest gas producers, Senegal, has
made recent additional natural gas discoveries but is in need of
Other LNG and gas-to-power initiatives. Total is also de- infrastructure to support LNG imports for gas-to-power proj-
veloping an LNG import terminal in neighboring Benin. The ects. With no new local gas production expected until 2025, Sen-
project includes an FSRU and an offshore pipeline connection egal has opportunities for LNG imports to ease its transition to
to existing and planned power plants in Maria Gléta. Accord- gas while also reducing the cost of its electricity generation. GP
ing to Total, Benin’s access to LNG will help the country meet
SHEM OIRERE is a freelance journalist based in Nairobi, Kenya.
growing domestic energy demand and add more natural gas to He has spent more than 10 yr covering various sectors of
the country’s energy mix, thereby reducing its carbon intensity. Africa’s economy, and has had numerous articles published
In Equatorial Guinea, a regasification plant to support the in several international publications and websites. He earned
a higher degree in journalism from the London School of
storage, transport and distribution of LNG across the country Journalism, and is also a member of the Association of
has been developed. The complex includes at least 12 bullet Business Executives (ABE).
NEW VERSION
InstruCalc •
10 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
SPECIAL FOCUS: TREATING SOLUTIONS
Gas processing plants offer a very different landscape than they boil. As a result, the mercaptans will distribute through-
the typical oil refinery, and the methods used to treat sulfur out various lighter hydrocarbon fractions. Since the strong
(S) compounds in the liquids must be adjusted accordingly. base, sodium hydroxide (NaOH or caustic), chemically reacts
When it comes to caustic treatment, one of the first configura- to remove acidic molecules H2S, CO2 and mercaptans, it can
tion decisions is whether to treat the C3–C10+ NGL stream be used to remove the mercaptans or convert them, depend-
in a single unit or to fractionate prior to caustic treatment. ing on the mercaptan type.
Treating the entire hydrocarbon stream may result in less Different methods of caustic treatment are used to remove
equipment, but fractionating can result in better selectivity for contaminants from the liquid stream, meeting S specifica-
product specifications. tions for pipeline mercaptan limits, downstream processing
Depending on the hydrocarbon source, LPG streams will requirements, process unit feed requirements, and propylene
have a vastly different distribution and concentration of mer- specifications.
captan molecules, which require more rigorous removal meth-
ods. Some contaminants, such as H2S, CO2, and mercaptan Alternatives in caustic treatment. Caustic treatment can
(RSH), can be removed from the LPG stream by caustic, re- be generalized into two different methods: extraction and
ferred to as extraction. Other contaminants, including COS, sweetening. Extraction is where mercaptans are removed from
CS2 or dimethyl sulfide (DMS), cannot be extracted from the the hydrocarbon to form low-S products. Fresh caustic can be
hydrocarbon with caustic alone. used to remove H2S, CO2, and mercaptan from these streams
Furthermore, the C5+ condensate streams also differ from with a once-through caustic contact scheme. However, once-
typical refinery naphtha streams. These streams can contain through caustic extraction has a high fresh caustic cost and a
very high levels of mercaptan that cannot be sufficiently re- high disposal cost. To offset this, most customers choose to
moved by caustic treatment. In this case, the mercaptan spe- catalytically regenerate the caustic, using air to convert and re-
cies must be converted to disulfides in an alkaline environment move the extracted mercaptans while returning the caustic to
over a fixed bed of catalyst to meet pipeline specifications. This the extractor with minimal S content. This regeneration meth-
process is referred to as “sweetening” because the mercaptans od has a higher capital cost than once-through caustic usage
are converted and the S content of the hydrocarbon does not but a much lower operational cost, due to the vast reduction
change. Other feed impurities will affect the design of the pre- in caustic consumption and disposal quantities.
treatment and metallurgy of the unit.
This article covers the most common impurities found in
TABLE 1. Boiling points of different compounds
caustic waste streams and how they affect the design and place-
ment of treating units in the midstream gas processing flow Boiling point, °F
scheme. These caustic treating systems also generate waste Compound Observed True
streams for disposal. As regulations on caustic disposal and S
oxides (SOX) emissions tighten, it is worth considering alter- Carbonyl sulfide –58
native options for disposal of spent caustic, disulfide oil and Propylene –54
spent air to meet ongoing and future changes in regulations. Propane –44
Most customers choose to catalytically acids, pipeline slip and flow enhancers, and upstream process-
ing chemicals. Treating this variety of contaminants can in-
regenerate caustic, using air to crease fresh caustic consumption and spent caustic disposal.
convert and remove the extracted Entire-stream extraction preferentially removes light mercap-
tans as opposed to larger mercaptans. Heavy mercaptans re-
mercaptans while returning the maining in the hydrocarbon would likely cause off-specifica-
caustic to the extractor with minimal tion product (see FIG. 2).
Conversely, entire-stream sweetening can better convert
sulfur content. This regeneration most of the heavy mercaptans to disulfide in the presence of
method has a higher capital cost than lower-strength caustic. This will cause a loss of light hydrocar-
once-through caustic usage but bon from spent air vented from the hydrocarbon. Additionally,
this process would need to have large fixed-bed reactors to ac-
a much lower operational cost. commodate the liquid flow and required space velocity.
Fractionation and individual treatment of narrowly cut frac-
The other type of caustic treatment is sweetening, which tions maximizes treating efficiency. Using the most effective
requires injection of air that dissolves into the hydrocarbon. caustic concentration for each treating objective optimizes se-
The oxygen from the dissolved air and the mercaptan already lectivity for mercaptan removal or conversion.
present in the hydrocarbon react in the presence of the solid Extraction designs are extremely flexible and can be opti-
bed catalyst and create disulfide that remains in the liquid hy- mized for mercaptan removal from lighter feeds [C1 and C2
drocarbon stream. This disulfide present in the product is less (gas phase) and C3 to C5 (liquid phase)] containing 2 wppm
corrosive than the mercaptan in the feed, despite no change in mercaptan S–20,000+ wppm mercaptan S in one or multiple
the total S. Caustic is easily separated from the hydrocarbon in tight cut extractors. This process can be designed to achieve
the reactor vessel while spent air is separated from the hydro- minimum reentry of S, reducing the S to less than 5 wppm S in
carbon, either in a degassing vessel in the process or in down- the product, and in some cases less than 1 wppm S. Since heavy
stream tankage, depending on the volume of spent air. Sweet- mercaptans are not in this cut, stronger caustic can be used,
ening can be used to treat naphtha (C5–C12) and to produce jet which increases the efficiency of the extractor, reduces caustic
fuel (C12–C20+) hydrocarbons. circulation rates and reduces vessel sizes.
Heavier condensate feeds (C5–C20+) will contain heavier
Pros and cons to fractionation prior to caustic treatment. mercaptans with concentrations ranging from 20 wppm S to
Treating the entire NGL stream without fractionation allows 2,000+ wppm S. These streams are more efficiently treated via
for minimum capital investment, but minimum efficiency. sweetening with weaker caustic strengths to increase heavy mer-
Using one caustic concentration results in poor selectivity captan solubility. Since the treated hydrocarbon has low vapor
for mercaptan reaction or removal. In addition, contamination pressure, minimum loss of hydrocarbon product occurs with
of caustic from pipeline chemicals can include interfering ions, the spent air vented from the hydrocarbon. Smaller-size fixed-
bed reactors can accommodate the liquid flow at the required
space velocity for the desired conversion. As a result, mercap-
tans are sweetened to disulfide to less than 5 wppm mercaptan
S for naphtha and less than 20 wppm mercaptan S for jet fuel.
C1SH Additionally, the product naphtha can be fractionated to
concentrate heavy disulfide S formed from sweetening the
mercaptans in the heavy cut, leaving a sweet, saleable light cut
of naphtha.
Equilibrium constant, log scale
C2SH
Typical and atypical impurities, and effect on design.
Several common impurities, such as CO2, COS, H2S and RSR,
can exist in the feed to the systems. Although it is expected
C3SH that entire-stream amine treating and fractionating of the
ethane cut would remove all the CO2 in the C3 and C4 cuts,
CO2 has been observed in these cuts at midstream facilities to
the concentration of 100 wppm–1,000 wppm. Caustic can be
C4SH used to completely remove CO2 and H2S, but it is consumed
irreversibly and would represent an operating cost in the range
of millions of dollars.
C5SH The preferred method of treatment at these high concentra-
tions is an amine absorber to reduce the concentration to less
than 25 wppm, followed by a caustic prewash upstream of the
Caustic concentration caustic treater. COS can be removed to less than 1 wppm by
pretreatment in a proprietary prewash design,a or during post-
FIG. 1. Mercaptan extraction curve.
treatment with a modified caustic blend.b
12 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
SPECIAL FOCUS: TREATING SOLUTIONS
Even though the typical impurities are present, atypical be reached in the extraction products by installation of a newly
impurities can affect the design of a complex. These atypical designed disulfide scrubber to drastically reduce the disulfide
impurities include oxygen, methanol, glycol, corrosion inhibi- present in the regenerated caustic. The disulfide scrubberd
tors, naphthenic acids and unusual mercaptan distributions. virtually eliminates reentry of S into the product streams by
Various corrosion inhibitors and slip enhancers will initiate counter-currently contacting the caustic stream with a hydro-
and stabilize emulsions between hydrocarbons and caustic. carbon stream, reducing the total product S by more than 95%
In addition, methanol can come from many upstream sources over a design without it.
(e.g., dehydrators) and form emulsions with caustic. Excessive The bottoms of the debutanizer can be processed through
methanol quantities should be removed with a water wash pri- a naphtha sweetening process.e This process treats C5–C12+ to
or to amine or caustic treating. Additionally, shared regenera- less than 5 wppm mercaptan S with minimum alkalinity re-
tion sections, discussed later in this article, transfer methanol quirements, which can then be used to dilute bitumen when
from one segregated stream to another. creating synthetic crudes.
Glycol and other oxygenates also must be considered. Another option after sweetening is fractionating the naph-
These impurities can cause acid corrosion in sweetening tha to produce a low-S C5 and C6 material.f Fractionating out
units. Two options exist to minimize the occurrence of acid the light C5 and C6 material concentrates the heavy disulfides
corrosion.c The first option is to construct the existing reac- created during the sweetening process in the bottoms product.
tors from 316L stainless steel. The second option is to apply
special epoxy to the carbon steel and concrete surfaces. This Methods for disposal of waste caustic and spent air.
creates an epoxy layer to protect the carbon steel of the reac- Caustic prewashing is necessary to completely remove the re-
tor. However, the epoxy will need to be replaced at every turn- sidual H2S and CO2 from the feed hydrocarbon. It is used up-
around of the unit. stream of the extraction process so that the feed is free of H2S
and CO2, which creates an S-laden caustic stream that must be
Integration of caustic treating with the complex. Differ- disposed of.
ent options exist to integrate caustic treating units into a grass- During the extraction process, mercaptan transfers from oil
roots or existing plant. One configuration may treat the entire to the aqueous caustic phase. The generated sodium mercap-
LPG stream with caustic extraction for use as power plant fuel, tide remains in the caustic, which is then separated from the hy-
low-S motor fuels, or sales to other processors. drocarbon phase. The sodium mercaptide is then catalytically
One example, shown in FIG. 2, places a C3 extraction pro- oxidized with injected air at almost ambient conditions to pro-
cess unit just after the depropanizer, thereby producing low-S duce disulfide, which separates from the caustic. This regenera-
propane required by a dehydrogenation process to produce tion process creates an S-laden spent air stream, caustic stream
propylene. Butanes from the debutanizer are treated in the and disulfide oil stream that must be disposed of (FIG. 3). The
C4 extraction process unit, which creates a low-S C4 cut for naphtha sweetening process also makes use of caustic and will
further fractionation by the deisobutanizer to generate low- produce a waste caustic stream that must be disposed of.
S isobutane and n-butane. The caustic is rich in mercaptides After the extraction and/or sweetening processes have
from both the C3 and C4 extraction systems and are sent to a been completed, several methods exist to dispose of the efflu-
common regeneration system to reduce capital cost. ent waste caustic, disulfide oil and spent air from the treating
Ultra-low-S specifications of less than 2 wppm total S can units. Each of these effluent streams have different disposal
Clean air
nViro
Low sulfur C2- MET Clean water
14 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
SPECIAL FOCUS: TREATING SOLUTIONS
Gas facilities use slug catcher ves- smaller than 2 µm); therefore, they were With myriad problems faced in the
sels to separate arriving gas and liquids. not capable of removing most of the Tanajib plant’s original slug catcher design,
These vessels are typically three-phase droplets, resulting in a very low separa- the following design considerations were
separators designed with vane-type mist tion efficiency and high liquid carryover. reviewed extensively, which resulted in the
eliminators. Mist eliminators are not When the operating pressure was above following stringent design requirements:
highly efficient in eliminating the liquid 870 psig/60 barg (typical for inlet facili- • Maintain the liquid carryover
carryover from the slug catchers, which ties), the droplets were usually smaller from the slug catchers equal to
will affect downstream separation pro- and the efficiency of separation was, or below 0.1 gal/MMsft3.
cesses, such as acid gas removal units. therefore, vulnerable. Furthermore, the • Maintain separation efficiency of
This article discusses enhancements plant’s processing capacity was often 98% for droplets larger than 10 µm.
created for the slug catcher’s internals reduced to minimize the accumulated
to achieve higher separation efficiency liquids in the downstream scrubber of Existing slug catcher designs. Two
and minimize the process design, siz- slug catchers, with frequent draining re- designs for slug catcher configurations
ing, material grade and future inspection quired for this scrubber. are discussed in the following sections.
and maintenance impact on downstream
processing units at Saudi Aramco’s Tana-
jib gas plant project in Saudi Arabia.
Historically, slug catcher vessels designed with Slug catcher 1 design. Slug catchers
in many existing facilities are designed
vane-type mist eliminators are not very effective with vane-type mist eliminators (FIG. 1).
The specified efficiency of 98% for drop-
in eliminating the liquid carryover from the slug lets larger than 10 µm and carryover of
liquid < 0.1 gal/MMsft3 required for the
catchers. The impurities in liquid carried over have Tanajib plant would not be achieved us-
a direct negative impact on downstream separation ing a vane pack mist eliminator, mainly
because of the very high operating pres-
processes, directly affecting the process design, sure of the slug catchers.
Historically, a vane pack (FIG. 2) was
sizing and material grade selection. found to be unsuitable for operating pres-
sure above 60 bar due to the fact that a
small liquid droplet size will be expected
due to the high pressure of the vessel.
The vane pack will not be capable of re-
moving most of these droplets, resulting
in very low separation efficiency and high
liquid carryover.
Slug catcher 2 design. In a second
slug catcher design configuration, the ex-
isting design had only a wire mesh mist
eliminator at the gas outlet nozzle. Due to
FIG. 3. Typical wire mesh mist eliminator configuration.
the high gas load factor, the mist elimina-
tor would not be able to achieve the re-
quired separation efficiency (FIG. 3).
FIG. 4. Illustration for a slug catcher with baffles and an axial mist eliminator.
16 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
SPECIAL FOCUS: TREATING SOLUTIONS
Process
Optimization
and Transition
Technologies
Why you need to attend:
IRPC-PROCESS.COM #IRPCPROCESS
SPECIAL FOCUS: TREATING TECHNOLOGY
Amine units used in gas and liquid hy- The plant was processing up to 90 evaluate the amine system because of its
drocarbon treating utilize various solvent MMsft3d of natural gas, but generally by- ability to accurately predict the actual
formulations. Some are amine-based, and passing up to 20 MMsft3d of gas to reduce operating conditions and performance of
some are not. These blended formula- operational costs while still meeting a 2.5% amine gas treating systems. The software
tions can be open-knowledge or propri- specification for CO2 content in the treat- package uses a mass transfer rate-based
etary formulas from a number of different ed gas. The plant normally operates with model for column calculations in the
amine suppliers. Usually, these formula- an amine circulation rate of 190 gal/min amine contactor and regenerator, as op-
tions are selected based on a review of the (GPM); the design circulation rate is 250 posed to equilibrium stage models used
plant design, feed acid gas content (CO2 GPM. The operator wanted to decrease in other simulators. The rate-based model
and H2S) and treated gas requirements. piperazine concentration in the amine sol- delivers a more accurate prediction of sys-
In the authors' experience, a considerable vent to reduce both solvent costs and rich tem operation and performance because it
number of amine units over-treat or, con- loading while staying just under the 2.5% takes into account the actual design of the
versely, are strained to meet specification specification. A conditional evaluation absorber column, as well as its internals.
with the solvent in place. Further evalua- was performed to understand if a potential Equilibrium stage models, by contrast,
tion of these units has revealed that modi- reformulation was feasible and, if so, what do not incorporate column design details
fications to the solvent formulation may conditions must be modified. into the simulation but instead rely on
result in more ideal treating conditions. The process was evaluated under sever- empirical data to estimate reaction kinet-
This article discusses the effects of al sets of conditions with varying MDEA/ ics. Equilibrium stage models use ideal
non-ideal amine formulations and the piperazine formulations to determine the stages for column calculations, which do
opportunities for reduced costs and im- lowest piperazine concentration required not account for actual tray or packed-
proved treating conditions made by re- to meet specifications. This article sum- bed designs and do not provide a predic-
formulating the solvent. A particular case marizes the evaluation work performed, tive output. The assumptions made for
is studied where a formulated methyl the results and interpretations, and the rec- column designs in an equilibrium stage
diethanolamine (MDEA) solvent was ommendations for solvent reformulation. model will be accurate only with em-
over-treating. An evaluation of several pirical data provided for that case, so any
formulations with reduced piperazine Process simulation overview. A rate- modification to the model input will not
concentration was performed to identify based simulator packagea was used to be simulated accurately without empirical
formulations that would potentially re-
duce solvent cost and rich CO2 loading.
data provided for the new conditions. The Deshmukh-Mather thermody- for mass and energy balances (but only
The objective of the evaluation was namic model is used by the software around ideal stages) and thermodynamic
to predict operational conditions and to predict vapor-liquid equilibrium be- phase equilibrium; generally, they do not
performance of the amine system with havior in the system, which is based on include an adequate aqueous electrolytic
different amine formulations. The rate- the Debye-Huckel theory regarding the model. Interfacial equilibrium, chemical
based model was necessary to provide ac- behavior of electrolytic solutions. This kinetic effects on mass transfer, and mass
curate predictions in the absence of any model predicts the complex interactions and heat transfer for components crossing
empirical data for potential new formula- involved in amine treating more accurate- interfaces are not accounted for.b
tions. Empirical data for the existing for- ly than conventional equilibrium models.
mulation was used to build a base model. However, a thermodynamic model Simulation design and conditions. To
That output matched with the empirical alone will not accurately predict amine simulate the system properly, normal and
data for the treated gas, thereby validat- system operation and performance. A worst-case operating conditions for the
ing the model accuracy. truly predictive model of amine system feed gas and lean amine were collected
operation and performance must ac- from historical data. The conditions used
Liquid film
count for five effects occurring in an ab- for this project are shown in TABLE 1.
Gas film Reaction film Diffusion film sorber or regenerator: The full composition of the feed gas
1. Mass and energy balance was also provided and used in the simula-
Pi kg
Pi* around phases on each tray tion. In addition, several design drawings
Bulk Bulk (or packing section) and schematics of the amine contactor
gas liquid
Physical 2. Thermodynamic phase were collected by plant personnel, and
solubility Chemical equilibrium (this model must be information on the sizing and tray design
Ci* solubility
activity-based for the components was incorporated into the simulation. Pa-
Ei
Ni CiL present and must include an rameters such as the number of trays, tray
Gas-liquid kio aqueous electrolyte model spacing, tower diameter, tray and valve
interface accounting for chemical reactions) type, and tray active area were used as in-
3. Interfacial equilibrium puts for the simulation.
Ni = kg (Pi – Pi*) = Ei kio (Ci* – CiL) between phases The simulation PFD is shown in
4. Chemical kinetics affecting FIG. 3. The feed gas composition and pro-
FIG. 2. Diagram representing the various mass transfer rates cess conditions were input into the feed
effects contributing to absorption in 5. Mass and heat transfer gas block; the amine formulation, amine
an amine contactor.b
for components and energy analysis results and process conditions
moving across interfaces. were input into the lean amine block; and
3: Treated gas FIG. 2 depicts several aspects involved the tower internals information was input
2
Lean amine with acid gas absorption in an amine ab- into the contactor block. The simulation
7 sorber. The absorption of acid gas compo- was then performed, and output data for
nents into the amine phase (Ni ) depends the rich amine, treated gas and gas-to-
Gas to pipeline
on the acid gas concentration in both pipeline streams were calculated. A feed
6: Bypass phases relative to the interface (Pi – Pi* gas flowrate prior to bypass of 90 MMs-
Contactor and Ci* – CiL), the film mass transfer coef- ft3d was used for the simulation with a by-
ficients (kg and ki0) and the enhancement pass flowrate of 10 MMsft3d–20 MMsft3d.
1 factor (Ei). The mass transfer coefficients
Feed gas 5 depend on equipment properties such as Results and interpretation. Several dif-
Rich amine tray/packing design, fluid properties, and ferent scenarios were simulated, using the
4 hydraulics. The enhancement factor de- program, by varying three key parameters:
pends on reaction kinetics and takes diffu- 1. Piperazine concentration
FIG. 3. Process flow diagram of the amine sion with reaction into account. 2. Feed gas flowrate
system evaluated. Equilibrium stage models account only 3. Lean amine flowrate.
The objective of the study was to deter-
TABLE 1. Process conditions used for simulation inputs mine if a combined gas-to-pipeline CO2
specification of 2.5% could be achieved
Parameter Value at lower piperazine concentrations, even
Feed gas flowrate, MMsft d 3
70 (normal)–80 (max.) at worst-case conditions. Piperazine con-
Feed gas pressure, psig 860
centrations were varied from 1.5%–3% in
different scenarios.
Feed gas temperature, °F 92–100 The lean amine flowrate was also var-
Feed gas CO2 concentration, mol% 4.69 ied in each scenario to potentially facili-
Lean amine flowrate, GPM 190 (normal)–220 (max.)
tate a more efficient treating process. In
some cases, an increase in amine flowrate
Lean amine temperature, °F 110–115 can reduce rich loadings to within recom-
20 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
SPECIAL FOCUS: TREATING TECHNOLOGY
mended guidelines (0.5 mol/mol) to pre- rich loading should reduce corrosion. At that at present conditions with the pres-
vent corrosion. In other cases, a decrease 80 MMsft3d, the gas-to-pipeline CO2 spec- ent solvent; this essentially showed that
in amine flowrate can reduce the reboiler ification was still met easily, but the rich this formulation was only feasible and ad-
duty while maintaining treated gas CO2 loading could not be reduced adequately, vantageous if the amine flowrate could be
content below specification. The feed gas even at the higher amine flowrate. This increased; otherwise it was more advanta-
flowrate to the amine absorber was varied evaluation showed that (1) some corro- geous to remain at 70 MMsft3d to the con-
between 70 MMsft3d and 80 MMsft3d, sion risk is present with the current solvent tactor with the present formulation.
but 70 MMsft3d was targeted to reduce at normal conditions that can be avoided At 2.5% piperazine, the same scenarios
rich loading and reboiler duty. by increasing amine circulation, and (2) were performed as with the 2.35% formu-
TABLE 2 shows the results of each sce- some corrosion risk will inevitably be pres- lation, and similar results were observed.
nario that was simulated. The data in- ent at 80 MMsft3d with the current solvent. The amine flowrate needed to be in-
cludes the variable inputs used for each The most aggressive scenario simulated creased for this formulation to work effec-
run, as well as the key performance indi- was using 1.5% piperazine. This simulation tively at the higher bypass rate and provide
cators predicted by the simulation output. was first performed at a 70-MMsft3d feed benefit through reduced rich loading.
Values in red were not within recommend- gas flow to the contactor and a 220-GPM With the assumption that the amine
ed guidelines or required specifications. lean amine flow to understand if enough flowrate could be increased, worst-case
The normal operational conditions CO2 could be removed at the higher by- scenarios with amine concentration
and current solvent were first simulated pass rate. The higher bypass rate allows for changes, feed gas temperature changes
to validate the accuracy of the model. The lower rich loading and is, in general, easier and lean amine temperature changes were
treated gas CO2 content predicted by the on the amine unit; therefore, lower bypass then simulated to understand the abil-
simulator matched very closely with that rates should be avoided when possible. ity of these solvent formulations to with-
measured at the plant during the same Since the CO2 content was well over 2.5% stand difficult conditions. Scenarios were
conditions (1.8%). The close match be- in the gas to pipeline, even at increased initially run at higher and lower amine
tween simulated and actual conditions amine flow, a higher level of piperazine concentrations, since amine is lost over
showed that the model was accurate and was determined to be necessary. At 2% pi- time and must be made up. It was found
that inputs could be modified to reliably perazine, the CO2 specification was again that treating efficiency decreases at lower
simulate real changes at the plant. barely missed in the gas to pipeline. concentrations, so several scenarios were
The normal operating conditions were At 2.35% piperazine, the CO2 specifica- performed at 52% water concentration.
also evaluated to understand if the unit tion was met in the gas to pipeline at 2.46%. TABLE 3 shows the same tests performed
could potentially be operated more ef- Other scenarios were then performed for as in TABLE 2, but with the total amine
fectively with the same solvent. The pre- this formulation. It was determined that at concentration reduced by 2% to consider
dicted rich loading at base conditions was the current amine flowrate, this formula- situations where amine is lost and has not
above the recommended maximum, so the tion would not meet the CO2 specification yet been made up. It was determined that
amine flowrate was increased to reduce it. unless the bypass flowrate was reduced. If both the 2.35% and 2.5% formulations
This resulted in higher CO2 removal and the bypass flowrate was reduced, however, were still able to meet CO2 specification at
increased reboiler duty, but the reduced the rich loading would increase to above reduced amine concentration.
TABLE 2. Simulation inputs and key performance indicators predicted by simulation output
Piperazine, MDEA, Water, Feed gas flow, Lean amine Weir load, Treated gas Combined gas Rich CO2 loading,
wt% wt% wt% MMsft3d flow, GPM GPM/ft CO2, mol% CO2, mol% mol/mol
*3% *47% 50% *70 *190 51–53 1.8% 2.46% 0.523
3% 47% 50% 80 190 51–53 1.87% 2.19% 0.583
3% 47% 50% 70 220 59–61 1.71% 2.39% 0.466
3% 47% 50% 80 220 59–61 1.82% 2.15% 0.513
2.5% 47.5% 50% 70 190 51–53 1.87% 2.51% 0.512
2.5% 47.5% 50% 80 190 51–52 1.93% 2.24% 0.573
2.5% 47.5% 50% 70 220 59–61 1.79% 2.45% 0.456
2.5% 47.5% 50% 80 220 59–61 1.88% 2.2% 0.504
2.35% 47.65% 50% 70 190 51–53 1.9% 2.53% 0.509
2.35% 47.65% 50% 80 190 51–52 1.95% 2.26% 0.57
2.35% 47.65% 50% 70 220 59–61 1.81% 2.46% 0.453
2.35% 47.65% 50% 80 220 59–61 1.9% 2.22% 0.501
2% 48% 50% 70 220 59–61 1.87% 2.51% 0.445
1.5% 48.5% 50% 70 220 59–61 1.95% 2.57% 0.434
*Normal operational conditions/current solvent
To further simulate non-ideal condi- the 2.5% formulation just barely misses The best formulation and conditions
tions for these formulations, the feed gas specification. At the same worst-case for further consideration were chosen by
and lean amine temperatures were also conditions, the 3% formulation still met finding a balance between adequate CO2
varied. Scenarios were run with these specification at 2.45% CO2. removal and rich loading. While several
changes at the higher bypass rate and in- scenarios met specification, most did so
creased amine flowrate, which would pro- Economics and recommendations. A at the edge of the high CO2 specification.
vide the most benefit for the unit overall. number of potential improvements were The simulations with 2.35%–2.5% pi-
These cases were also performed at re- discovered in the evaluation results. It was perazine yielded the most ideal results in
duced amine concentration to consider primarily recommended to increase the that both loading and CO2 content could
the worst-case scenario. amine flowrate to ≥ 220 GPM if corrosion be maintained comfortably below recom-
TABLE 4 shows that at elevated feed gas in the rich amine circuit was of concern. mended maximums if the amine flowrate
temperature and reduced amine concen- This would reduce rich loading to levels could be increased. Some worst-case con-
tration, the 2.35% formulation would not of acceptable corrosion risk. In addition, ditions emerged where the 2.35%–2.5%
meet CO2 specification (unless the by- the weir load on the trays was low at nor- formulations did not meet specification at
pass flowrate is increased). The 2.5% for- mal operating conditions, just above the the higher, 20-MMsft3d bypass rate; how-
mulation was able to meet specification minimum guideline of 50 GPM/ft. Spray ever, in those situations, the specification
at elevated feed gas temperature, but only flow can occur at low weir loads, leading can be met by slightly reducing the bypass
if the amine temperature is maintained; to reduced treating efficiency and solvent rate. It follows, then, that more efficient
at elevated amine temperature and the carryover, but an increase in amine circu- treating would be possible with a reduced
other non-ideal conditions specified, lation could resolve the issue. piperazine concentration during normal
TABLE 3. Simulation inputs and key performance indicators predicted by simulation (low amine concentration)
Piperazine, MDEA, Water, Feed gas flow, Lean amine Weir load, Treated gas Combined gas Rich CO2 loading,
wt% wt% wt% MMsft3d flow, GPM GPM/ft CO2, mol% CO2, mol% mol/mol
2.88% 45.12% 52% *70 *190 51–53 1.82% 2.47% 0.543
*(3%) *(47%)
2.88% 45.12% 52% 80 190 51–52 1.89% 2.2% 0.605
2.88% 45.12% 52% 70 220 59–61 1.72% 2.39% 0.485
2.88% 45.12% 52% 80 220 59–61 1.83% 2.15% 0.534
2.4% 45.6% 52% 70 190 51–53 1.88% 2.52% 0.532
*(2.5%) *(47.5%)
2.4% 45.6% 52% 80 190 51–52 1.94% 2.25% 0.597
2.4% 45.6% 52% 70 220 59–61 1.79% 2.45% 0.475
2.4% 45.6% 52% 80 220 59–61 1.89% 2.21% 0.524
2.26% 45.74% 52% 70 190 51–53 1.9% 2.53% 0.53
*(2.35%) *(47.65%)
2.26% 45.74% 52% 80 190 51–52 1.96% 2.27% 0.592
2.26% 45.74% 52% 70 220 59–61 1.81% 2.47% 0.472
2.26% 45.74% 52% 80 220 59–61 1.91% 2.23% 0.521
*Concentration at 50% water
TABLE 4. Simulation inputs and key performance indicators predicted by simulation (non-ideal conditions)
Piperazine, MDEA, Water, Amine Feed gas Weir load, Treated gas Combined gas Rich CO2 loading,
wt% wt% wt% temperature, °F temperature, °F GPM/ft CO2, mol% CO2, mol% mol/mol
2.88% 45.12% 52% 115 100 59–61 1.79% 2.45% 0.474
*(3%) *(47%)
2.4% 45.6% 52% 115 92 59–61 1.81% 2.46% 0.473
*(2.5%) *(47.5%)
2.4% 45.6% 52% 110 100 59–61 1.85% 2.49% 0.466
2.4% 45.6% 52% 115 100 59–61 1.86% 2.5% 0.464
2.26% 45.74% 52% 115 92 59–61 1.83% 2.48% 0.47
*(2.35%) *(47.65%)
2.26% 45.74% 52% 110 100 59–61 1.87% 2.51% 0.463
2.26% 45.74% 52% 115 100 59–61 1.89% 2.52% 0.461
*All simulations performed at 70 MMsft3d to the contactor and 220 GPM lean amine
22 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
SPECIAL FOCUS: TREATING TECHNOLOGY
conditions, but at non-ideal conditions prepared based on thorough simulation in chemical and biological engineering from the
University of Colorado at Boulder.
the current formulation would work more and evaluation of the amine unit design
effectively (to a certain extent). and operating conditions—is imple- CODY RIDGE is a Lead Process
It was recommended that a formula- mented less frequently. Plant personnel Engineer at Amine Optimization
tion with approximately 2.35% pipera- should reassess, at least yearly, whether Company and a Chemical Engineer
from Texas Tech University in
zine be used if the amine flowrate could or not such an investigation should be Lubbock, Texas. He is responsible
be increased to 220 GPM. If the amine performed with their current amine sol- for field engineering and
flowrate could not be increased, then a vent, as a proper evaluation may reveal technology development with
3% formulation would be necessary. The opportunities for reduced solvent costs, Amine Optimization. Mr. Ridge has worked as an
operator and process engineer in the Permian Basin.
economic implication of reducing the operating costs, energy usage, corrosion
content of piperazine in MDEA solvents and other cost-effective performance DAVID ENGEL has more than
based on present market values was, on improvements. GP 25 yr of industrial experience in a
average, $0.1/lb–$0.15/lb ($0.9/gal– variety of chemical engineering
NOTES and chemistry areas. He is the
$1.3/gal) of solvent for each 1% reduc- inventor in 21 US patents and the
a
ProTreat process simulator package
tion in piperazine. Reducing the pipera- b
Based on ProTreat technical literature author of a number of technical
zine concentration in the present solvent and scientific papers. Dr. Engel
from 3% to 2.35% implied a total savings SCOTT WILLIAMS is a Process has developed business and technology for
Engineer at Amine Optimization. Eastman Kodak, Eli Lilly, Pentair, General Electric
of $0.57/gal–$0.86/gal of MDEA sol- He has industry experience in a and Sulphur Experts worldwide. He has specialized in
vent, or $2,860–$4,290 for a 5,000-gal number of projects in oil and gas, chemical engineering, process chemistry, optimization
inventory. (Note: the data is based on ac- petrochemical, chemical and and contaminant removal technologies. He is the
cess to market price values). water treatment applications. Managing Director of Nexo Solutions Companies.
As part of the Amine Optimization He holds a BS degree in industrial chemistry, a MS
This study and others have shown engineering group, Mr. Williams is responsible degree in chemistry and a PhD in organic chemistry.
that plants can benefit from an evalua- for technical design and solutions development in He is also Six Sigma and Project Management
tion of their amine solvent in place, es- engineering and technology applications. He also certified. Dr. Engel is a Committee Member of the
pecially as feed gas compositions change provides support for analytical and specialized American Filtration Society, Southwest Region; and a
service projects. His recent work has been focused member of the Gas Processors Association Technical
over time. Several standard solvents are on amine unit contamination control, process stability Section M, in addition to a member of the boards of
sold to plants, but an ideal formulation— and energy reduction. Mr. Williams holds a BS degree directors at several companies.
INCLUDES:
• Natural Gas Liquids
• Midstream gas processing
• LNG, NGL, GTL processing technologies
• Pipelines and storage
• Transport, shipping, and distribution
H2-TECHSOLUTIONS.COM #H2TECH
TURBOMACHINERY
the turbine over a 2-yr period. This ap- authors noted that the gas turbine was represents 8% of the time of good opera-
proach consisted of observing a single stopped 15 times in 2 yr for general tion, which indicates that the repair team
gas turbine during a certain time of op- cleaning, largely as a result of climate ex- is doing well since it has reached a maxi-
eration and under real conditions of use, posure. The three graphs in FIG. 2 show mum of 17 hr. Finally, in FIG. 2C, a com-
and listing all the failures that arose and other preliminary statistics. parison of uptime with repair time shows
the information relating to these failures In FIG. 2A, it is shown that the great- that the average uptime is decreasing as
[e.g., time between failure (TBF), time est number of malfunctions occurred the repair time increases.
to repair (TTR) and downtime (DT) between 0 hr and 420.4 hr. Six recorded As a preliminary reading of these sta-
due to breakdowns]. This list provided malfunctions prove that the turbine is tistics, the overall system health of this
the basic data to quantify the reliabil- going through a difficult stage in terms gas turbine is moving closer to the dan-
ity of the gas turbine under study. The of good work. In FIG. 2B, the repair time ger zone.
A
100 7.46 on a balanced relationship between pre-
loss
100 1.05
Total 2.27 89.22 dictive and preventive maintenance. The
time Availability Quality
Availability
80 losses losses
Speed
P primary goal of reliability analysis is to
losses OEE
analyze the behavior of systems based on
loss
Percentage %
60
their failure rate data. These assessments
OEE = A × P × Q
40
most commonly used method is reliabil-
Quality
20
Run time
Useful
time
0
was studied in the context of the Weibull
0 1 2 3 4 5 6 distribution, based on the history of in-
terventions for a limited period of time
FIG. 3. Cascade diagram of the OEE of the studied gas turbine.
(14 mos).
26 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
TURBOMACHINERY
where η is the scale parameter (dimension FIG. 4. Method for reliability analysis of the studied gas turbine.
of t), γ is the location parameter (dimen-
sion of t) and β is the shape parameter (di-
mensionless). TABLE 1. History of interventions
To calculate the aforementioned pa-
No. (N) of DN × max. =
rameters, the data must be prepared by intervention MTBF, hr N Σni F(i) F(t) [F(i) – F(t)]
first determining the pairs (ti, Fi) by the
average (median) ranks, as shown in Eq. 5: 1 93 1 1 0.0454 0.0993 0.0539
TABLE 3. Maintainability results for the studied turbine increasing curve (FIG. 5), which comple-
ments the unity of the probability and
Mean TTR, hr 5 12 3 6 4 17 9 2 7 ensures that the system is not repaired on
M(t) 0.58 0.864 0.394 0.632 0.485 0.943 0.776 0.28 0.688 the interval (0, t). From the previous re-
sults, the system is shown to be maintain-
1.00 able at 63.2% at t = 6 hr.
The availability is shown to be de-
creasing, and the availability is a reflec-
tion of the reliability and maintainabil-
0.75 5.99 Maintainabilty, M(t): 0.632 ity. The change in availability is a direct
result of the changes in reliability and
Maintainability, M(t)
• Check the reliability of other tem subject to discrete external load,” Journal of Diagnostic Laboratory at the University of Djelfa.
Reliability Engineering and System Safety, Vol. 193, Dr. Hafaifa additionally leads the Gas Turbine Joint
systems with links to turbine August 2019. Research Team in support of several international
equipment. GP 8
Glovatskii, O.-Y., “Operating experience and reli- research projects in the industrial sector. His research
ability assessment of elements of pumping stations,” interests include control in industrial systems,
Journal of Hydrotechnical Construction, Vol. 23, Iss. 9, reliability engineering, fault detection and isolation
ACKNOWLEDGMENTS September 1989. in industrial processes and intelligent systems based
This work was supported by the Applied 9
Zhejun, G., “Estimation of mixed Weibull distribu- on fuzzy logic and neural networks. He has also
Automation and Industrial Diagnostics Laboratory tion parameters using the SCEM-UA algorithm: published a number of papers in industry journals.
with the Gas Turbine Joint Research Team at the Application and comparison with MLE in auto-
University of Djelfa in Algeria. The authors express motive reliability analysis,” Journal of Reliability MOSTEFA TELHA is a Professor in
their sincere thanks to the staff of the Menia-Ghardaia Engineering & System Safety, Vol. 91, Iss. 8, August the mechanical department of the
power station for their kind hospitality during the real- 2006. science and technology faculty at
ization of this work. Djelfa University in Algeria. He is
preparing his PhD in fluid
NADJI HADROUG is an Associate
mechanics at the University of
LITERATURE CITED Professor with the faculty of science
Laghouat in Algeria. He is an
1
Zohair, A., A. Hafaifa and S. Abudura, “Operational and technology at Djelfa University
engineer in industrial maintenance, and his work
reliability analysis applied to a gas turbine based in Algeria, as well as a member of
focuses on energy systems and the maintenance
on three-parameter Weibull distribution,” Mechanics, Djelfa University’s Applied
and diagnostics of mechanical systems.
Vol. 21, Iss. 3, 2015. Automation and Industrial
2
Saranyasoontorn, K. and L. Manuel, “Efficient mod- Diagnostics Laboratory. He received ABDELHAMID IRATNI is an
els for wind turbine extreme loads using inverse his PhD from the University of Ziane Achour in Associate Professor in control
reliability,” Journal of Wind Engineering and Industrial Djelfa, Algeria. His work focuses on the development engineering and automation
Aerodynamics, Vol. 92, April 2004. of new methods and tools in fault-tolerant control for systems in the electrical
3
Meslameni, W., “Optimization of maintenance times industrial systems, and his research interests include engineering department of the
of a gas turbine,” International Journal of Applied the monitoring and control of industrial gas turbine faculty of science and technology
Research and Technology, Vol. 1, 2018. systems. Dr. Hadroug is the author and co-author of at the University of Bordj Bou
4
Shen, M. H. H., “Reliability assessment of high cycle several publications and conference papers. Arréridj in Algeria. He has served as the Chair of
fatigue design of gas turbine blades using the proba- the scientific committee of the electrical engineering
bilistic Goodman diagram,” International Journal of AHMED HAFAIFA is a Full department since 2019, after serving as the head
Fatigue, Vol. 21, Iss. 7, August 1999. Professor in industrial automation of the bachelor’s and master’s programs since 2015.
5
Naikan, V. N. A. and S. Kapur, “Reliability modelling and diagnosis and reliability Dr. Iratni received his BSc and MSc degrees in
and analysis of automobile engine oil,” Journal of engineering at the University electrical engineering from the University of
Automobile Engineering, Vol. 220, December 2004. of Djelfa in Algeria, where he Boumerdes in Algeria in 1999 and 2003, respectively.
6
Rausand, M., “Reliability centered maintenance,” conducts research work in applied He received his PhD from the University of Sétif in
Journal of Reliability Engineering and System Safety, automation and industrial Algeria in 2013. His research areas include nonlinear
Vol. 60, 1998. diagnostics. He also serves as the Dean of the filtering, estimation and control, biomedical and
7
Zhang, J., Y. Zhao and M. Xiaobing, “Reliability Science and Technology Department and the bioprocess engineering, and automation and
modeling methods for load-sharing k-out-of-n sys- Director of the Applied Automation and Industrial reliability of industrial systems.
ORDER TODAY!
PRICE: $1,695
HydrocarbonProcessing.com
BACK TO BASICS
Natural gas liquids (NGL) are a group of hydrocarbons, ies to make motor gasoline, as petrochemical feedstock and as
sometimes referred to as purity NGL, that includes ethane, pro- diluent for heavy crudes.
pane, normal butane (n-butane), isobutane and natural gasoline As can be inferred, NGL markets are rather convoluted as
(pentanes plus). NGL made their inroad into fuel markets in the they entail five different, volatile markets sharing a common
1910s, when the issue of gasoline loss while stored under normal source of supply. Moreover, the transportation and storage of
conditions was addressed, leading to a way to bottle liquid gas. NGL are expensive as they require relatively high pressure and
NGL are produced predominantly in gas processing plants low temperature to be maintained in the liquid state.
and, to a lesser extent, in refineries as homogeneous liquid blends Since NGL are mixtures with low density and low viscosity,
of purity NGL. The NGL extracted from natural gas make up the they can be used as diluent for heavy crude oil so that it can be
so-called Y-grade NGL from which ethane, propane, n-butane, efficiently transported via pipeline. When the crude oil is worth
iso-butane and pentanes plus are separated by fractionation in a more than the weighted average value of NGL as gas, the conden-
distillation unit. A subset of NGL is liquid petroleum gas (LPG). sate can be used for crude spiking. In doing so, not only are the
LPG is a blend of propane, n-butane and i-butane. NGL valued as crude, but also the crude may be upgraded due to
All natural gases contain NGL. Although the quantity of liq- the increase of API gravity and sold at a higher price per barrel.
uids is not always great enough to be economically recovered, Recovery of NGL might be economically attractive if they
there are cases where NGL extraction is desired either to meet are worth more than their value in sales gas; the relative value of
sales specifications (minimum and maximum high-heating- liquids against gas is given by the frac spread—i.e., the difference
value requirements) or to produce transportable gas—i.e., a gas between the NGL market prices and the value of retaining them
that can be piped without hydrocarbon condensation along the in the natural gas. Eq. 1 explains how the frac spread is calculated:
pipeline (dewpoint control).
This article provides an overview of technology for the recov- Frac spread = $l – $g × HHVCi × Fi ÷ 1,000,000 (1)
ery of NGL, as well as a brief discussion on the economic reasons
that justify the extraction of NGL from natural gas and provide where:
the foundation for different processing facility configurations. $l = Price of a purity NGL, $/gal
$g = Price of natural gas, $/MMBtu
Fundamentals of NGL economics. NGL have a wide variety HHVCi = High heating value of the purity NGL, Btu/sft3
of applications, including feedstock for petrochemical produc-
tion, as fuel for vehicles and as specialized fuels for space heating
and cooking.
Among individual NGL, ethane has the largest share of NGL
field production, followed by propane. Together, they make up
90% of the NGL barrel. Ethane is largely used to produce ethyl- T = –30°C/
ene, the monomer used to produce polyethylene, ethylene ox- –50°C
ide, ethylbenzene and dichloromethane, which are raw materials Residue gas Natural gas Turbo-
booster expander
for the production of a wide variety of everyday objects.
Propane is the precursor of polypropylene, which is used to T = –90°C/–100°C
produce plastics, resins, rubbers and other materials. It is also Raw gas P up to 30 barg
used as a residential and commercial heating fuel, as a drying C3 chiller
agent for crops, as a ripening agent for fruit, and so on.
N-Butane is used to produce butadiene, the key component
in synthetic rubber. Butane is also used during winter as a fuel NGL
Reboiler
additive in motor gasoline. I-Butane is mainly used in refineries
as feedstock for the alkylation process to produce alkylate. It is Demethanizer
also used in refrigerators, as fuel, and as propellant in cooking
FIG. 1. Turboexpander process flow diagram.
spray and hairspray. Pentane is used as a blending fuel in refiner-
Gas Processing & LNG |JANUARY/FEBRUARY 2021 31
BACK TO BASICS
The cold two-phase stream from the expander heads to the Residue gas
de(m)ethanizer, where the light components dissolved in the
condensate are stripped out according to the operating mode
Raw gas C3 chiller or
(C2 extraction or rejection mode). The Y-grade NGL are ex- J-T valve
Booster* Fuel gas
tracted from the bottom of the de(m)ethanizer and routed to Gas/gas
the fractionation section of the facility, while the residue gas is exchanger
LTS separator
compressed in the compression end of the turboexpander.
LTS
Since only a fraction of the mechanical work produced by separator
Flash
the expander can be transferred to the compressor, the pressure drum
NGL
of the residue gas is significantly lower than that of the raw gas. Rich-lean
To feed the downstream pipeline, a booster compressor must be Lean MEG heat exchanger Glycol
installed to restore the pipeline pressure. The booster compres- regeneration
sor is a major consumer of electric power in the process plant. *Only for JT process
Rich MEG
Some processing technology (e.g., Fluor’s TCHAP) has been
specifically developed to minimize this recompression cost. FIG. 4. Simplified PFD of a mechanical refrigeration plant.
It is worth mentioning that in cryogenic service, the most
widely used heat transfer equipment is the brazed aluminum
heat exchanger (BAHE) or the more recent printed circuit heat
exchanger (PCHE). A BAHE is a stack of alternating layers Fired heater
of parting sheets and heat transfer fins. Each layer is provided
with inlet and outlet ports for the streams and is sealed along
the edge by means of a side bar, as shown in FIG. 2. The block
Coalescer
of layers is bounded by cap sheets at the top and bottom and is Knockout drum
provided with headers and nozzles for the entrances and exits Raw gas Filter NGL
of process flows. Adsorbers Water
Air cooler
BAHEs feature a large surface area per unit volume, high
heat transfer, a minimum temperature approach as low as 1°C,
Normal operation Residue gas
and the ability to handle multiple process streams. Compared Regeneration circuit Final filter
with a shell-and-tube heat exchanger, a BAHE can be 80%–85% Cooling circuit
smaller and lighter. BAHEs are generally made of aluminum.
Since aluminum is prone to mercury corrosion, the use of BA- FIG. 5. Process scheme for adsorption on silica gel.
HEs require a mercury removal unit in the pretreatment section
of the central processing facility. Moreover, aluminum is a weak
material, making BAHEs vulnerable to thermal shocks during absorber tower between the turboexpander and the deethanizer
startup and shutdown operations. (FIG. 3A). A cold, ethane-rich stream coming from the deethanizer
PCHEs are compact heat exchangers without gaskets, joints is pumped to the top of the absorber, and the bulk of the C3+ rich
or welds. The core of the heat transfer is obtained by means of stream is extracted from the bottom. Generally, the dual column
diffusional bonding, which is a solid state joining process. A also entails the dual reflux concept for the bulk removal of NGL
number of platelets are stacked in a pile, pressed together, heat- in the absorber. In this arrangement, the deethanizer overhead
ed at 50%–70% of the material melting point and then allowed gas is partially condensed, and the liquid is used to reflux both the
to rest so that the material grains can grow and fill the gaps be- deethanizer and the absorber towers. In the dual-column, dual-re-
tween the surfaces. In doing so, a monolithic block of core heat flux configuration, as much as 99% of propane can be recovered.
exchange is obtained. The channels for fluid flow are obtained Split-feed gas subcooled process. A slipstream (FIG.3B),
by chemical etching or pressing. typically 15%–30% of the main stream exiting the separators of
the cooling train, is subcooled against the cold residue gas and
Designs to accommodate market volatility. When design- refluxed to the de(m)ethanizer tower (in the case of a single-col-
ing central processing facilities for NGL production, processing umn arrangement) or to the absorber tower.
facilities must be made to accommodate both the ethane extrac- Tailgate gas recycle. A slipstream of compressed tailgate gas
tion and ethane rejection modes to tune the plant operation to is cooled in the feed multi-stream plate-fin exchanger, expanded
market volatility. Rejecting ethane entails a partial loss of pro- and fed to the top of the de(m)ethanizer. This arrangement im-
pane, which is generally more valuable as a liquid than as natural proves the C2+ extraction.
gas. Therefore, a design goal would be to maximize the propane Once the Y-grade NGL have been extracted from natural gas,
recovery while ensuring the maximum ethane rejection. each purity NGL is obtained by separation of the Y-grade in the
The different ways in which these objectives have been fractionation end of the production chain. The fractionation is
pursued has given rise to different distinctive features of the accomplished by means of a direct sequencing operation of dis-
various processing technologies. The fundamental concepts tillation, in which the lightest component is taken overhead in
underlying these different ways can be traced to the basic con- each column. Ethane is first separated from the liquid blend, and
cepts outlined in the following sections. then propane, n-butane, iso-butane and natural gasoline are sepa-
Dual column. A dual column consists of the installation of an rated in sequence.
Gas Processing & LNG | JANUARY/FEBRUARY 2021 33
BACK TO BASICS
TABLE 2. NGL extraction technologies frigeration is achieved by exploiting the cooling effect caused by
the pressure drop across a throttling valve ( JT control valve).
Commercial name Processes Technology provider The process scheme is similar to that of mechanical refrigera-
Cryogenic turboexpander technology tion. Indeed, in the JT process scheme the mechanical refrigera-
CRYOMAX DCP, MRE, Flex-e TechnipFMC tor is replaced with the JT valve.
High Propane HPA, NGL MAX, McDermott/Lummus
As the pressure let down is an isoenthalpic and isoenthrop-
Recovery NGL PRO Technologies ic transformation, the final temperature achievable with this
Gas Subcooled GSP, SCORE Ortloff
process is lower with respect to the external refrigeration, but
Process, Single Column higher compared to the expander technology. Therefore, the JT
Overhead Recycle process allows the recovery of a greater quantity of NGL than
LPG/C3+ and NGL/ Cry-Plus, ROC, Linde the external refrigeration with propane, but a smaller quantity
C2+ recovery and GSP, RSV than the expander.
STANDARD line For tight pressure difference between the raw gas and the
Cryo-Gas TCHAP, TRAP, DDP, Fluor residue gas (the majority of cases), the JT process will eventually
VRAP, SARD, ERGR require the installation of a booster compressor to restore the gas
Cryogenic process without expander pressure, with sizeable OPEX penalization.
Absorption on solvent AET NGL Rec. Advanced Extraction
Adsorption on silica gel. Adsorption on silica gel is a sepa-
Technology ration process (FIG. 5) based on surface chemistry, more spe-
Propane refrigeration Open art
cifically on physi-sorption consisting of hydrocarbon and water
cycle bonding on hydroxyl groups distributed throughout the surface
Iso pressure open IPOR McDermott/Lummus
of the silica gel. Being an amorphous material with mesopores of
refrigeration Technologies approximately 20 Å, the pores of the silica gel are also the locus
of C6+ capillary condensation; this characteristic enhances the
LPG recovery LPG Plus Black & Veatch
efficiency of natural gasoline separation.
Joule–Thomson Open art By increasing the temperature, the interactions between hy-
Adsorption processes drocarbons and silica gel loosen. After all the sorbent hydroxyl
Generic adsorption Open art groups have been engaged and the bed has become saturated,
process the adsorbent can be regenerated by heating. Overall, the re-
Advanced adsorption ADAPT Siirtec Nigi through moval process with silica gel is a dynamic process, where the ad-
process DNV GL sorption stage at temperatures lower than 38°C for a single bed is
followed by a regeneration stage at 230°C–270°C.
Absorption on solvent. In this process, a stream of gasoline A silica gel plant consists of at least one fixed-bed column in
(C5+) is used as solvent for extracting NGL in a regenerative ab- adsorption mode, one in regeneration mode and one in cooling
sorption process where the raw gas is contacted counter-currently mode. The continuity of the operation is achieved through se-
with a cold lean solvent. The rich solvent drawn from the bottom quencing of the columns.
of the absorber is regenerated in a fractionation tower and re- The ADAPT technology utilizes the heat pulse technique,
cycled back in the absorber. The refrigeration duties required to which consists of heating up only a portion of the bed and then
reflux the columns are provided by a propane refrigeration cycle. exploiting the heat accumulated in this portion to heat up the
Arguably, this process is less efficient and more expensive remaining part of the bed. This technology makes possible the
relative to the more competitive and simpler expander plant. regeneration and cooling in a single column. In doing so, the
This explains why few industrial NGL plants are based on this equipment count and the related, costly cyclic valves and head-
technology. ers are considerably reduced, with sizeable CAPEX savings.
Mechanical refrigeration. In this process, NGL are partially Recently, the ADAPT technology has been extensively im-
liquefied by cooling natural gas to a temperature as low as –37°C plemented for dewpoint control for international transmission
to –40°C against an evaporating refrigerant fluid, normally pro- of more than 125 Bm3/yr of natural gas through long, subma-
pane, in a kettle-type heat exchanger of a closed-loop refrigeration rine pipelines without intermediate recompression facilities.
system. Part of the refrigeration duty is recovered in the gas-gas This process requires deep NGL removal to avoid slug forma-
heat exchanger (GGHE), where the raw gas is cooled against the tion in the pipelines. GP
residue gas from the low-temperature separator (LTS) (FIG. 4).
LORENZO MICUCCI is a Senior Director at Siirtec Nigi SpA.
During the cooling step, the gas crosses the hydrate forma- He has more than 30 yr of experience in the engineering and
tion envelope and the risk of heat transfer equipment clogging contracting industry, most of which have been spent in the
increases; therefore, a hydrate inhibitor must be injected up- natural gas sector. In 2001, he joined Siirtec Nigi in Milan, where
stream of the GGHE. The inhibitor can be either methanol or he directed the process design and operations department and
the research and development department. During his time
glycol—mainly monoethylene glycol (MEG). The former is a as R&D head, three patents have been granted to Siirtec Nigi,
relatively volatile chemical; therefore, MEG is generally pre- two of which have been implemented on an industrial scale. At present, he is the
ferred to methanol. The mechanical refrigeration is, in practice, Senior Director of the technology and marketing departments. Mr. Micucci also
worked for Saipem (Snamprogetti) as a Plant Designer for integrated gasification
an isobaric process; therefore, it can be implemented only if the combined cycle and gas-to-liquids plants. He holds an MS degree in chemical
operating conditions fall beneath the cricondenbar. engineering from the University of Bologna in Italy and is enrolled as a Qualified
Joule–Thomson (JT) process. In this process, the gas re- Engineer in the Register of Milan Order of Engineers.
34 JANUARY/FEBRUARY 2021 | GasProcessingNews.com
NEW IN GAS PROCESSING TECHNOLOGY
A. BLUME, Editor-in-Chief
NORTHEAST U.S.,
EASTERN CANADA
KOREA
YB Jeon
ADVERTISER INDEX
Merrie Lynch +82 (2) 755-3774
+1 (617) 594-4943 Korea@GulfEnergyInfo.com Filtration Technology Corporation .......... 5
Merrie.Lynch@GulfEnergyInfo.com RUSSIA, FSU
Catherine Watkins, Publisher
Lilia Fedotova
Gas Processing & LNG ................................23
+1 (713) 520-4421 OUTSIDE NORTH AMERICA
Catherine.Watkins@ +7 (495) 628-10-33 Gulf Energy Infrastructure........................... 2
HydrocarbonProcessing.com AFRICA Lilia.Fedotova@GulfEnergyInfo.com
www.GasProcessingNews.com Dele Olaoye UK, SCANDINAVIA, IRELAND, Gulf Software ................................................ 10
NORTH AMERICA +1 (713) 240-4447 MIDDLE EAST H2Tech.........................................................7, 24
Africa@GulfEnergyInfo.com Sam Wibrew
NORTH HOUSTON, NORTH BRAZIL +44 (0) 208 133 7678 HPI Market Data 2021 ..................................30
CENTRAL TEXAS, FORT WORTH Evan Sponagle Sam.Wibrew@GulfEnergyInfo.com
Jim Watkins Phone: +55 (21) 2512-2741 Neil Levett
IRPC Process ..................................................18
+1 (713) 525-4632
Jim.Watkins@GulfEnergyInfo.com
Mobile: +55 (21) 99925-3398 +44 (0) 7774 290522 Merichem........................................................36
Evan.Sponagle@GulfEnergyInfo.com Neil.Levett@GulfEnergyInfo.com
WEST HOUSTON, CENTRAL CHINA, HONG KONG Pipeline Opportunities Conference .......... 8
TEXAS, NEW MEXICO CORPORATE, FULL ACCESS
Iris Yuen
Brett Stephen SUBSCRIPTION AND
China: +86 13802701367 This index and procedure for securing additional information are
+1 (713) 525-4660 Hong Kong: +852 69185500 CLASSIFIED SALES
Brett.Stephen@GulfEnergyInfo.com provided as a service to advertisers and a convenience to our readers.
China@GulfEnergyInfo.com J’Nette Davis-Nichols
Gulf Energy Information is not responsible for omissions or errors.
DALLAS, MIDWEST/CENTRAL U.S WESTERN EUROPE +1 (713) 520-4426
Josh Mayer Jnette.Davis-Nichols@
Hamilton Pearman
+1 (972) 816-6745 GulfEnergyInfo.com
+33 608 310 575
Josh.Mayer@GulfEnergyInfo.com Hamilton.Pearman@GulfEnergyInfo.com
WESTERN U.S., INDIA
BRITISH COLOMBIA Manav Kanwar
Rick Ayer +91 (22) 2837 7070
+1 (949) 366-9089 India@GulfEnergyInfo.com
Rick.Ayer@GulfEnergyInfo.com
ITALY, EASTERN EUROPE
SOUTHWEST HOUSTON, Riccardo R.C. Laureri
SOUTH TEXAS, UPPER MIDWEST, Office: +39 02 2362500
ALBERTA Mobile: +39 335 6962477
Nathan Frank Riccardo.Laureri@GulfEnergyInfo.com
+1 (713) 525-4608
JAPAN
Nathan.Frank@GulfEnergyInfo.com
Yoshinori Ikeda
SOUTHEAST HOUSTON, +81 (3) 3661-6138
GULF COAST & SOUTHEAST U.S Japan@GulfEnergyInfo.com
Austin Milburn
+1 (713) 525-4626
Austin.Milburn@GulfEnergyInfo.com