Gas 2025 Issue

gas 2025

PTQ supplement

BTX FROM PYGAS

LNG OUTLOOK

AI-BASED STRATEGIES

OFFGAS MONETISATION

UPDATE YOUR SUBSCRIPTION REGISTER FOR A PRINT OR DIGITAL ISSUE TO MAKE SURE YOU CONTINUE RECEIVING A REGULAR ISSUE OF PTQ ANY QUESTIONS? REGISTER CONTACT US CLICK HERE

gas ptq PETROLEUM TECHNOLOGY QUARTERLY Editor Rene Gonzalez editor@petroleumtechnology.com tel: +1 713 449 5817 Managing Editor Rachel Storry rachel.storry@emap.com Editorial Assistant Lisa Harrison lisa.harrison@emap.com Graphics Peter Harper Business Development Director Paul Mason sales@petroleumtechnology.com tel: +44 7841 699431 Managing Director Richard Watts richard.watts@emap.com Circulation Fran Havard circulation@petroleumtechnology.com

3 Gas-based products dominate low-carbon fuels growth Rene Gonzalez Editor, PTQ

7 Corrosion mitigation of amine units using MEA for deep CO₂ removal: Part 2 David B Engel, Scott Williams and Cody Ridge Nexo Solutions 13 Maximising value from refinery off-gases Wolf Spaether, Holli Garret, Kristina Morgan and Felix Schulz Clariant 22 Effect of maldistribution on tail gas treating unit absorber performance G Simon A Weiland, Prashanth Chandran and Ralph H Weiland Optimized Gas Treating, Inc. 27 Advances in distillation processes for BTX aromatics production David Kockler Consultant

EMAP, 10th Floor, Southern House, Wellesley Grove, Croydon CR0 1XG tel +44 208 253 8695

33 Resolving FeS contamination and poor filtration in a liquid treating system Nicholas Brownrigg Howard Energy Partners Orbie Harris IV Transcend Solutions, LLC 37 Dynamic simulation of an LNG plant fuel gas system Harry Z Ha, Cole Beattie and Javeed Mohammed Fluor Canada Ltd.

Cover Shell’s liquefied natural gas (LNG) regasification terminal in Gibraltar, switching from diesel-fuelled power

generation to cleaner-burning natural gas. Photo courtesy of Shell International Limited.

©2025. The entire content of this publication is protected by copyright full details of which are available from the publishers. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means – electronic, mechanical, photocopying, recording or otherwise – without the prior permission of the copyright owner. The opinions and views expressed by the authors in this publication are not necessarily those of the editor or publisher and while every care has been taken in the preparation of all material included in Petroleum Technology Quarterly and its supplements the publisher cannot be held responsible for any statements, opinions or views or for any inaccuracies.

Our analyzers make tough SRU work a little easier WE’RE COMMITTED TO YOUR EFFICIENCY We know how hard it is to manage sulfur recovery unit (SRU) processes, and the levels of skill, concentration, and dedication your team needs. That’s why we make sure you don’t have to worry about your analyzers, too. We’ve been designing industry-standard SRU analyzers for decades, focusing on reliability, longevity, accuracy, robust design, and ease of use. We make analyzers for every part of the sulfur removal process – from SRU feed gas to the measurement of stack emissions, and everything in-between - so you get one convenient source for unparalleled engineering and support.

AMETEKPI.COM/SRU

EDITORIAL COMMENT

Gas-based products dominate low-carbon fuels growth

Market demand, infrastructure, and technological innovation are making gas-based projects more scalable, cost-competitive, and sustainable

Rene Gonzalez Editor, PTQ

I nvestment in gas-based products, primarily lique- fied natural gas ( LNG), continues expanding with the upstream discovery and development of new gas fields in the Middle East, Africa, US (shale-based gas production). Closely connected to midstream pipeline/terminal infra- structure, a fleet of very large LNG carriers, real-time inte - grated fleet monitoring systems, and other enablers have made LNG the dominant gas-based product. Diversification into decarbonised power sources such as solar and wind continues falling short of meeting demand for new power and fuel sources, while growth in the EV market has yet to meet expectations. Meanwhile, emerging LNG consumer markets at a mega scale (for example, India) and a small scale (for example, Caribbean Basin) favour the construction of LNG liquefaction facilities along the US Gulf Coast, The Middle East, Africa, and elsewhere. While a detailed review of LNG and other gas-based technology is beyond the scope of this article, suffice it to say that several liquefaction processes have been developed, with the main differences seen in the type of refrigeration cycles used. These processes can be broadly classified into two groups: mixed refrigerant processes and cascade liquefaction processes (using pure components as refrigerants). LNG is seen as a bridge fuel in the shift away from coal and oil, providing lower emissions and supporting global climate goals. Rising energy consumption in China, India, and Southeast Asia supports increased LNG imports as these regions transition from coal to cleaner energy sources, although coal consumption is still reportedly high. Industries and power plants are increasingly using LNG as a reliable energy source due to its efficiency and lower car - bon footprint. Countries are looking to diversify energy supply sources to reduce dependency from specific regions (for example, Europe reducing reliance on Russian gas), leveraging flex - ible LNG supply solutions based on large-scale and small- scale LNG (ssLNG) trains to liquefy natural gas down to -162°C for storage and transportation. The size of modu - lar ssLNG units makes them ideal for sites near cities and industrial parks to fuel electric generators and power data centres to meet the demands of AI workload. Another example more closely related to the refining

industry is the Stockholm ssLNG terminal, which can sup- ply LNG to the neighbouring Nynas Nynäshamn refinery. From this source, the refinery can generate the hydrogen it needs for its hydroprocessing units. According to recent projections, the switch to natural gas from naphtha will reduce the facility’s CO₂ emissions by more than 57,000 tonnes per year. As LNG production scales, costs continue to decrease, making LNG more competitive against coal, oil, and renewa- bles in certain markets. As with AI-centric data centres, new types of consumers are reducing cost factors. For example, the shipping and maritime industries are shifting towards Improved liquefaction and regasification technologies based on new processing techniques enhance LNG production efficiency, lowering capital and operating costs LNG as a fuel source to comply with IMO (International Maritime Organization) emissions regulations. Other emerging LNG consumers include expanding mining operations (such as lithium and cobalt for EVs and other rare earth metals), shifting remote temporary power requirements from diesel-fuelled generators to natural gas- fired generators in the 100 kW to 1.0 MW range. Elsewhere, higher natural gas liquid (NGL) volumes are providing feed - stock for steam cracker-based olefins production. Other developments in reducing LNG industry costs include float - ing LNG (FLNG) plants. These modular plants lower project costs, reduce environmental footprint, and bring new sup- ply to market. Other enablers involve larger, more efficient LNG carriers (tankers) where advancements in ship design, such as Q-Max and Q-Flex LNG carriers, reduce transpor- tation costs and improve efficiency (see Figure 1 ). Long-term vs short-term opportunities Improved liquefaction and regasification technologies based on new processing techniques enhance LNG pro- duction efficiency, lowering capital and operating costs.

Gas 2025

www.digitalrefining.com

Figure 1 LNG carriers navigating the high seas are equipped with the latest international safety protocols Courtesy of Shell International Limited

This includes advancements in LNG storage and cryogenic technology. For example, more efficient storage solutions, including floating storage regasification units (FSRUs), make LNG accessible to more markets. In support of these developments, AI-driven predictive maintenance and auto- mation improve plant uptime, reduce operational costs, and enhance safety. In the long term, the potential to blend hydrogen with LNG could position it as a long-term transition fuel towards a low-carbon future. However, this involves significant chal - lenges. For starters, LNG is stored at around -162°C, while hydrogen liquefies at -253°C, making it difficult to store and transport them together efficiently. Hydrogen can cause embrittlement in steel pipelines and infrastructure, poten- tially leading to safety concerns. Hydrogen has a higher flame speed and lower energy density than natural gas, requiring modifications to turbines, burners, and engines. At present, there are around 70 LNG terminals worldwide. Small-scale facilities such as the one in Nynäshamn are still the exception but offer a significant benefit, which is now boosting demand in other regions such as the Caribbean basin. The LNG market is projected to grow at a compound annual growth rate (CAGR) of more than 4.9% between 2025-2030, with most capacity coming from mega-LNG facilities. However, many of the niche margin opportunities are coming from ssLNG operations. Unlike mega facilities, modularised ssLNG facilities in the 0.5 to 2 million metric tonnes per annum (mmtpa) range can be brought online relatively quickly, reducing overall cost and complexity, such as the Eagle LNG Partners project in Jacksonville, Florida. Long-term ssLNG market prospects are favourable as island nations and smaller countries transition towards LNG

as a fuel option. Consequently, flexible US cargoes may prove to be attractive options. Currently, the mega-LNG owners are more exposed to price sensitivity regarding long-term con- tracts and other tolling agreements, making shut-ins of one or more LNG trains more likely in the short term. For exam- ple, the lack of a reliable power grid on the Caribbean islands predicates the benefits of modular LNG-powered generators in the 50 MW or less range. Demand for power India has become a major influencer in the global LNG market as it develops the infrastructure for importing LNG. Recent press reports noted that India’s gas-fired power generation doubled in the spring of 2024 to 8.9 billion kilo - watt-hours (kWh) compared with the same period in 2023. More than 75% of India’s power generation was from coal in 2023, while gas-fired plants have accounted for only about 2% in recent years, largely because of the high cost of gas relative to coal, which is why Indian refiners are mak - ing efforts to increase gas recovery from different refinery offgas and flaring operations. India’s domestically produced gas is largely being used for fertiliser production and cooking fuel in cities. Indian LNG imports are forecast to reach more than 28 million metric tons in 2025, up from 22.1 million tons in 2023, according to Independent Commodity Intelligence Services (ICIS). This is why other gas sources, such as refinery off - gas (ROG) recovery and the opportunity for cogeneration (from offgas), are important considerations going forwards. While renewables are growing in importance worldwide and will continue to do so, LNG is unique in its diversity of applications. For example, Vietnam is focusing on imported

Gas 2025

www.digitalrefining.com

Impact on olefins product yields

Ethane cracking: Yields ~80% ethylene, minimal propylene and heavier products Propane cracking: Yields ~40-45% ethylene, ~15% propylene Butane cracking: Yields ~35-40% ethylene, ~20-25% propylene Naphtha cracking: More balanced yields of ethylene (~30-35%), propylene (~15-20%), and aromatics

Table 1

LNG with plans to build 15 new LNG-fired power plants by 2035 with a combined capacity of more than 22 GW. According to Shell’s LNG Outlook 2024 , global demand is expected to increase by more than 50% by 2040. This will be driven largely by the move away from higher-emitting fossil fuels. In addition to heightened energy security, this leads to economic development. New facilities Two major LNG projects in the US, including Plaquemines LNG and Corpus Christi Stage III, hit major milestones in December, and feed (as deliveries) reached a new record high in January. Comments from the new US Administration seem to indicate a preference for LNG capacity. The first six blocks of Plaquemines LNG have Federal Energy Regulatory Commission (FERC) approval to introduce hazardous fluids, and feed gas deliveries have already surpassed 1.3 Bcf/d. Meanwhile, press releases indicate Corpus Christi Stage III achieved its first LNG on 30 December 2024, and Cheniere expects Train 1 of the project to reach ‘substan - tial completion’ at the end of 1Q25, with production antici - pated to ramp soon thereafter. Looking to the end of 2025, the 2.05 Bcf/d Golden Pass LNG project is expected to start up in December, according to the guidance given in Exxon’s 3Q24 earnings call. However, the company noted there is a risk that the project could slip into early 2026. Other gas-based resources In addition to LNG, both condensate (API gravity: 55-65°) and NGLs (C₂H₆, C₃H₈, and C₄H10) play a crucial role in the energy industry. Condensate production has increased sig - nificantly in recent years, driven by the growth of uncon - ventional gas production. It has become an important component of the global oil market, with its price often linked to benchmark crude oil prices. For example, in the US, wells operated by a half-dozen exploration and pro - duction companies (E&Ps) in eastern Ohio’s Utica Shale are now churning out more than 100 Mb/d of superlight crude oil (aka condensate), more than twice as much as three years ago, according to a report by H. Carr in 2025. NGLs, on the other hand, have a well-established market and are traded separately from crude oil. They are in high demand for various industrial and residential applications, making them an essential part of the energy supply chain, including steam cracking to olefins. Increasing the use of NGLs in steam cracking is a trend driven by the growing availability of NGLs from shale gas production. This shift impacts the production of ethylene, propylene, and other key petrochemicals.

As with LNG, the rapid expansion of shale gas resources has led to increased availability of NGLs (ethane, propane, and butane), often cheaper feedstocks than naphtha, leading to lower production costs. Ethane and propane cracking pro - duces a higher percentage of ethylene compared to naphtha (see Table 1 ). However, ethane cracking produces almost no propylene, requiring additional on-purpose propylene pro - duction (for example, propane dehydrogenation, PDH). However, as with LNG, there are challenges when shift - ing to NGL feedstocks in steam cracker operations, such as accurately estimating the investment needed in fractiona - tion, transportation, and storage to handle increased NGL use in mixed feed steam crackers. A detailed discussion on mass transfer and product separation concerns is beyond the scope of this discussion. Suffice it to say that process adjustments and steam cracker designs must be optimised for different feedstocks to balance heat integration and byproduct handling (see Figure 2 ). Many new plants, such as new and recent projects along the US Gulf Coast and the Middle East, are designed spe - cifically for ethane cracking, while other steam crackers are being modified for mixed feeds to balance ethylene and propylene production. Shifting to more NGLs in steam cracking requires adjustments in process design, reaction conditions, and downstream processing, including feed - stock handling and preheating, where the fractionation units in NGL service must be separated into pure ethane, propane, or butane before cracking. The fractionation units (cryogenic distillation) are critical for this, and the ethane and propane require preheating to vapourise fully before entering the cracking furnace. Special heat exchangers and vapourisers also must be used. Just as challenging is the cracking furnace design, which involves high-temperature pyrolysis. NGL cracking occurs at 750-900°C (1,380-1,650°F) in the presence of steam to prevent coking. To maximise ethylene yield and minimise coke formation, the gas stays in the reactor for a short res - idence time of only 0.1-0.5 seconds. Other complexities with the NGL involve radiant coil design, where ethane and propane require different coil designs from naphtha. For example, ethane cracking uses shorter, high-radiance coils to optimise heat transfer. Additional processing challenges Figure 2 World-scale steam crackers benefit from the abil - ity to process mixed feeds Courtesy of BASF SE

Gas 2025

www.digitalrefining.com

when feeding NGL to a steam cracker, such as steam ratio and coke reduction, product quenching, and the use of dual feedstock crackers that can switch between NGLs and naphtha to optimise production, will be discussed in future issues of PTQ . Associated process gas With crude oil production, there are opportunities to mon- etise co-produced gas or associated process gas (APG). Technically, a number of options exist for handling APG, including preparing it as fuel in various forms, such as dried pipeline gas or liquefied petroleum gas (LPG), and export - ing it via pipeline. It can then be re-injected for later recov- ery, such as processing it as LNG or LPG and exporting it via tankers for converting it to petrochemical feedstock. Modular or portable gas-powered engines driving elec- trical generators (Gensets) can turn APG into power for onsite electricity, eliminating the cost of diesel deliveries to remote areas and reducing CO₂ emissions that otherwise might result from diesel fuel consumption or previous vent- ing or flaring practices. As with other gas sources, there are challenges to its commercialisation, such as the need for gas desulphurisation at relatively high H₂S concentra - tions, but cleaning this fuel makes it valuable, especially in remote areas with limited fuel sources. In addition, the waste heat from the Gensets can also be used for onsite heating purposes. Stranded gas While natural gas or well-gas-driven gensets may cost more to rent than a diesel generator, long-term fuel cost savings make natural gas generators more economical when also factoring in Tier 4 emissions compliance require- ments. According to feedback from industry experts, the diesel engine is normally 40% of the genset’s component cost (engine + generator), so when the cost of the engine increases by 80 to 100% (for technology needed to meet more stringent emissions compliance), the overall cost for diesel generators becomes significantly more expensive than gas generators, invalidating the industry paradigm that diesel generators are typically the lowest capital option. In addition, a payback of greater than 10-to-1 may be real- ised with discounted gas as diesel prices increase in response to the IMO 2020 global 0.5% sulphur cap for marine bunker fuels (against the current 3.5% limit). The sulphur cap will require blending diesel into bunker fuels to meet the 0.5% sulphur spec that cannot be met by the heavier distillates and high sulphur fuel oils (HSFOs) currently in use. Incentives Multiple proposed US projects, including Venture Global’s CP2 LNG facility, Commonwealth LNG, Rio Grande Phase 2, Port Arthur Phase 2, and Lake Charles LNG, may be able to make significant strides this year towards reaching the market. Additionally, the new US Administration provides a tailwind for facilities with non-free trade agreement (FTA) approvals that are set to expire within the next four years, such as Cameron LNG Phase 2, Woodside Louisiana LNG, and Texas LNG. Collectively, these eight facilities total

around 15 Bcf/d of nameplate capacity and would nearly double current US LNG capacity. Although not necessary to move forwards with an LNG project, Department of Energy (DOE) non-FTA authorisation is important for new LNG projects, as it allows access to critical demand centres, making it more attractive for securing offtakers and financ - ing and ultimately paving a path for a facility to achieve a positive final investment decision (FID). Elsewhere, the first train of LNG Canada, a 1.8 Bcf/d facility, is expected to be fully in service in mid-2025. The project will be the first major export project in the country. Additionally, Western LNG, with the Nisga’a Nation and Rockies LNG, is looking to achieve a positive FID on its Ksi Lisims LNG project in 2025, a proposed 1.5 Bcf/d facility, after it received a significant investment from Blackstone Energy Transition Partners. BTU Analytics had previously identified the growing Canadian LNG industry as a risk to US gas markets in the August 2024 edition of the Gas Basis Outlook . However, the same can be said of other projects in the Western Hemisphere. According to a recent report from Energy Capital and Power, several significant LNG projects are in develop - ment in Africa. The first phase of the 2.3 million tons LNG Greater Tortue Ahmeyim (GTA) project between Senegal and Mauritania is nearing completion and commissioning. Other major projects are on track in other parts of Africa. Gas-based projects in The Middle East, such as the 1.0 mmtpa Marsa LNG project located within SOHAR Port and Freezone, will provide maritime energy solutions in the Middle East with a cleaner fuel alternative for global ship- ping while facilitating Oman’s economic diversification. Another project of high interest due to its unique status of making it one of the world’s lowest carbon intensity LNG plants is the Ruwais LNG project in the United Arab Emirates. It will include the construction of two liquefaction trains, each with a capacity of 4.8 mmtpa. This new capac- ity, totalling 9.6 mmtpa, will more than double ADNOC’s current UAE LNG production, bringing it to approximately 15 mmtpa as the company expands its international LNG portfolio. Located in Al Ruwais Industrial City in the Al Dhafra region of Abu Dhabi, the project will be the first LNG export facility in the Middle East and North Africa (MENA) region to operate on clean power. Vision According to the most recent Shell LNG Outlook , global LNG demand for liquefied natural gas (LNG) is forecast to rise by around 60% by 2040, and more than 170 million tonnes of new LNG supply are set to be available by 2030. A few of the drivers for the expanding gas industry have been discussed, not the least of which is leveraging clean gas to heat and power homes, businesses, and industries throughout the world. Take, for example, the development of natural gas fields in and around relatively remote Trinidad and Tobago and conversion to LNG and NGLs through liq- uefaction facilities on those Caribbean islands. The reve- nues earned contribute to GDP in that area, allowing for a higher standard of living for many in that region while reducing the need for coal, diesel, and heavy oils.

Gas 2025

www.digitalrefining.com

Corrosion mitigation of amine units using MEA for deep CO 2 removal:Part 2

Amines used in gas treating affect reboiler and regenerator stainless and carbon steel components, influenced by temperatures, chlorides, acid gas loadings, and other factors

David B Engel, Scott Williams and Cody Ridge Nexo Solutions

I n Part 1, published in PTQ Q1 2025, various corrosion chal- lenges faced by amine units using MEA solvent during their operation were discussed, along with basic information related to the background of amine solvents and corrosion events. In Part 2, the content is developed in greater depth in terms of the chemical aspects and their process implications. H 2 S and CO 2 removal by amines Monoethanolamine (MEA) is a primary amine and the strongest amine when compared to secondary (DEA) or ter- tiary (MDEA) amines. MEA has substituted one single etha- nol group (CH 2 -CH 2 -OH), leaving two hydrogens attached to the nitrogen in the molecule (see Figure 1 ). All gas treating amines (primary, secondary, or tertiary) react instantane- ously with hydrogen sulphide (H 2 S), using their loan pair of electrons over the nitrogen. However, they all react differ- ently towards carbon dioxide (CO 2) . The CO 2 replaces the hydrogen attached to the nitrogen in MEA. Thus, the presence of hydrogen in the MEA chem- ical structure means that there are two active sites for CO 2 reaction. This makes MEA an attractive molecule for both H 2 S and deep (very low) CO 2 removal in key applications. Typically, CO 2 and H 2 S can be removed to values less than 5 ppmV. The loan pair of electrons over the nitrogen in MEA is very active for reactions, especially with steel and corrosion, which limits solvent strength. Amine solvents, in general, have low corrosivity and have historically been used as corrosion inhibitors in multiple appli- cations. Nevertheless, when amines are subjected to acid gas loading, the allowable strength must be limited based on how aggressively the amines and their salts attack the metal surface. Testing work presented at a gas conference in 1991 showed the relative corrosion tendencies of the three types of alkanolamines in relation to their concentrations (see Figure 2 ). Typical acceptable corrosion rates for amine units are <5 mils/yr.

common CO 2 removal process using an amine unit, such as at the gas plant in this article, CO 2 corrosion can occur in any zone where the CO 2 partial pressure is high, tem- peratures are elevated, or solvent velocities are high. Any combination of two to three of these factors often results in severe corrosion. Since the H 2 S partial pressure is very low, there will be minimal protective iron sulphide film on the walls of the unit, leaving the CO 2 to form pits that could pass right through the walls of the unit equipment. High CO 2 contents com- bined with warm/hot contactor temperatures generally cause CO 2 attack on the contactor walls (via carbonic acid attack caused by the CO 2 dissolving in the condensed water on the vessel walls), which manifests as pitting corrosion at the hot zones in the contactor tower. Corrosion in amine units using MEA solvents is primarily focused where the contactor tower maximum temperature occurs (bulge), which often times is near the middle section of the contactor given the low absorption rates at the bot- tom of the column. The predicted temperature profile in the contactor, as presented in Figure 3 , shows the temperature bulge in the mid-section of the column. Amine regeneration and corrosivity Even though the lean amine loading has not dramatically exceeded the recommended values, as the rich amine load- ing has, there are still significant issues in the regeneration of

15 10 5 20 25 30 35

In this testing work, in order to get meas- urable corrosion rates, the testing was done at elevated tempera- tures in a continuous CO 2 atmosphere. In a

Loan pair of electrons

MEA molecular structure H CH CH OH H N

Solvent type

Figure 2 Relative corrosion tendencies of alkanolamines ( Depart, LRGCC 1991 ) Hot skin corrosion test, CO 2 atmos- phere, carbon steel, seven-day test @210°F (99°C)

Figure 1 MEA molecular structure

Gas 2025

www.digitalrefining.com

Hence, regeneration was not taking place properly. The low temperature out the top of the regenerator is a result of the loss of steam at the top of the column section. The steam generated in the reboiler has three objectives:  Heat the feed solvent to the reboiler temperature (sen- sible heat load).  Break the reaction bond between the CO 2 /H 2 S and the MEA.  Provide enough energy for a reflux flow between 7-10% of the main circulation flow rate. For the regenerator heat load, the energy required to heat the solvent is increased when the feed temperature is cooler than the recommended 195°F minimum. According to plant data, the rich amine feed temperature was only 160°F (71ºC), which means the steam has to heat up the solvent by 90°F instead of 45-55°F in a typical amine unit. The extra steam wasted in heating up the solvent reduces the amount of steam available to regenerate the CO 2 and H 2 S from the rich amine solvent. This means the regenerator runs out of steam before steam reaches the top of the regenerator, so rather than stripping acid gases out in the top of the regenerator (ideal situation), rich amine travels deeper down the regenerator and even into the reboiler before the amine is finally regen - erated. This is one of the most common causes of regener- ator corrosion. H 2 S is a slightly weaker acid gas than CO 2 (carbonic acid) , so it is more easily regenerated higher up the regenerator. Excessive CO2 in the hottest part of the regenerator and reboiler leads to excessive corrosion all the way through the hot lean piping until the amine solvent is cooled in the lean/ rich (L/R) exchanger. As the simulation graphs show (see Figure 4 ), CO 2 is still being regenerated in the bottom of the regenerator, whereas most H 2 S is removed by the time the amine solvent flows into the reboiler. This is a situation with a very high corrosion potential.

Contactor t emperature pro f ile

Vapour Liquid

Temperature (Fahrenheit)

Figure 3 Predicted absorber temperature bulge

the amine that will cause corrosion of both carbon steel and stainless steel. Regeneration of the amine solvent is achieved by counter-currently reacting the solvent with steam flowing up the regenerator tower. MEA is a very strong primary amine, so it does not want to give up its reacted acid gas easily. It takes significant steam energy to regenerate rich MEA. One can tell that the unit has generated sufficient steam to properly regenerate the rich amine solvent by checking the regenerator overhead temperature. For MEA solvents, this temperature must be 225-235°F (107-113ºC). The actual value was only 192°F (89ºC).

Regenerator : Hydrogen sulphide in vapour phase

Regenerator: Carbon dioxide in vapour phase

0E+0

5E+2

1E+3

2E+3 1.5E+3 Concentration (ppmw) 2.5E+3

3E+3

3.5E+3

4E+3

0E+0

1E+5

2E+5

4E+5 3E+5 Concentration (ppmw) 5E+5

6E+5

7E+5

8E+5

Figure 4 H 2 S and CO 2 desorption curves in the regenerator column. H 2 S (left) CO 2 (right)

Gas 2025

www.digitalrefining.com

0.6

Fe (ppmw) Rich loading Lean loading

0.5

0.4

0.3

Jan 2020 New pump

June 2020 New L/R exchanger

0.2

Jan 2018 New contractor

0.1

Figure 5 Iron content vs lean and rich loadings

Regenerator and reboiler corrosion To mitigate corrosion in the regenerator, regeneration of 95% of the rich amine must happen before entering the reboiler. A second rule is to reduce reboiler and vapour return line corrosion to maintain the CO 2 content of the reboiler return below 1 mol%. According to the simulation, only 89% of the rich amine is regenerated prior to the reboiler. The vapour return line contains 1.2-2.5 mol% CO 2 . The following operating param- eters need to change to regenerate more in the column, thus reducing the corrosion potential:  The L/R heat exchanger needs to be upgraded to achieve a rich outlet temperature of at least 190°F.  Have sufficient hot oil flow/temperature to the reboiler to achieve a regenerator overhead temperature of at least 215°F (102ºC) (preferred >220°F, 104ºC). The simulated water reflux flow rates are near 0.5 GPM or less. The expected flow rate should be closer to 3-5 GPM for a properly regenerated MEA solvent. By raising the regener - ator overhead temperature close to the recommended val- ues, the water reflux rates will increase accordingly. Corrosion in amine units can be initially detected via:  A change in solution colour, clarity, and visible solids.  Levels of dissolved iron, manganese, chromium, and Pure MEA solvent is clear; however, in a process with H2 S, corrosion of steel surfaces forms iron sulphide, which tends to turn the solvent pale yellow, dark yellow, green, brown, and finally to a black colour as corrosion continues unabated. This takes place as the iron content builds from virtually zero ppmw to greater than 50 ppmw. If there is not enough H 2 S to form an appreciable passivation iron sulphide (FeS) layer on the unit piping/equipment, the majority of the iron will be in the form of iron carbonate (Fe 2 CO 3 ). Colours build from clear to yellow, to gold, to brown to black as Fe 2 CO 3 levels increase. Iron and manganese are typically representative of car- bon steel corrosion. Iron levels represent current corrosion nickel in the amine solvent and analysis.  Levels of certain suspended solids.

rates, while manganese levels represent historic corrosion. Chromium and nickel are typically present in solution when stainless steel corrodes. Usually, chromium dissolves in the solution and can be detected in a laboratory analysis, while nickel precipitates as a solid that is filterable. Iron content below 5 ppmw is acceptable, whereas chromium content greater than 1 ppmw is problematic. Figure 5 provides a plot of the iron in solution compared to the lean and rich amine loadings. Corrosion at high temperatures If new equipment was installed, an increase in soluble iron would be consistent with the high initial corrosion rates of fresh steel. Usually, after the first few weeks, corrosion rates tend to decline as a protective passivation film forms on the inner metal surfaces. An event in December 2019 was the initial onset of iron in the solution; although relatively small, it was still noticeable against the historic values below the detectable limits of the test. Given the anomaly in the rich amine loading at the same period, another change occurred in either the operating mode or equipment. Our assumption is a reduction in rates to allow for the equipment changes. Overall, the soluble iron levels were acceptably low. However, in early January, there was a massive increase in the soluble iron content. This coincided with the installation of a new pump off the regenerator column. This replacement caused corrosion because of the elevated temperatures and insufficient H2 S or CO 2 present to form protective films. Since corrosion is a chemical reaction and all chemical reactions proceed faster and more aggressively as the solution heats up (corrosion rates can often double for every 16ºF increase in temperature), reboiler and lower regenerator column corrosion will always be high. Figure 6 provides a plot of chromium levels. As with iron, the chromium content made a similar increase around the same time frame. With chromium levels this high, it gener- ally indicates a significant failure of stainless steel compo - nents. Generally, the locations of interest would be the trays

Gas 2025

www.digitalrefining.com

250

Cr (ppmw) Fe (ppmw)

200

150

100

Figure 6 Iron and chromium content in the amine solvent

in the regenerator tower, reboiler bundles, stainless steel piping, or stainless peripherals (pump internals or thermal probes). This appears to be caused by the pump installed at the regenerator column where several solids were circu- lated. The pump was installed because of corrosion at the outlet bucket of the regenerator. Harsh pitting While iron content has continued to decline up to August, the chromium content rose after attaining an equilibrium value of around 40 ppmw. High stainless steel corrosion without a correspondingly high carbon steel corrosion tendency is not normal for amine units. The reason stainless steel is chosen for amine unit metallurgy is because of corrosion resistance (H 2 S and CO 2 corrosion plus velocity-related corrosion). Corrosion resistance is a result of a thin (about 5 nanometers) oxide layer on the steel’s surface. The passivation layer forms because of the chromium added to the stainless steel surface. There are several reasons stainless steel will corrode in amine units. The first is aggressive chloride (Cl-) content in

the solvent. Depending on the supplier, the maximum chlo- ride content is 250-500 ppmw. As Figure 7 shows, the chlo- ride levels were often within the maximum range. Chlorides cause chloride stress corrosion cracking (CSCC) of stainless steel, sometimes described as ‘harsh pitting’. Carbon steel is often too soft to crack in the presence of chlorides. Another reason for stainless steel corrosion would be the presence of large amounts of foulant (possibly by corrosion products), causing a loss of local oxygen and subsequent loss of the protective oxide layer. Recently installed probes that corroded rapidly had no solids covering them when extracted. The probe likely suffered from distinct corrosion mechanisms (CO 2 pitting and CO 2 cavitation from CO 2 release). High chloride levels It is also possible that a stainless steel corrosion probe sitting in stagnant liquid, where the chloride content in the solvent was enough to disrupt the protective oxide film, resulted in pitting of the steel. The probe showed distinct rounded pits often associated with CO 2 attack. The high lean amine

500

450

400

350

300

250

200

150

100

Figure 7 Chloride ions in solution

Gas 2025

www.digitalrefining.com

loading possibly contributed to the attack along with chlo- rides. It is also possible that the probe was made with a grade of stainless steel with poor chloride resistance (410SS, 304SS, for example). 316L should be the stainless steel of choice for facilities operating with elevated chloride levels. The amine unit was operating with a high lean and rich amine loading. High amine loadings corrode carbon steel rapidly. Normally, to return the rich amine loading within guidelines, the unit should increase the solvent circulation rate. However, the reboiler limitations could not regener- ate the current low solvent flow rates, and any increase in recirculation rates would require an increase in reboiler duty (25% increase in recirculation rates translates to an increase of 25% in reboiler duty). Another option to lower the rich amine loading was a higher solvent strength. However, a 20 wt% maximum concentration MEA is recommended (solution corrosivity increases as the strength increases). The regeneration of the rich amine appears to be occur- ring too low location-wise in the regenerator column. The prime reason for this is a low rich amine feed temperature entering the regenerator (rich outlet temperature from the L/R exchanger). The L/R exchanger needs to be upgraded to attain a minimum outlet rich amine temperature of 195°F (91ºC). Combined strategies The amine unit needs to target a regenerator overhead temperature of at least 215°F (102ºC), or corrosion in the hot lean amine circuit of the unit will continue to take place. The elevated stainless corrosion in the unit is not a normal occurrence when signs of carbon steel corrosion are low. High chloride levels in the solvent are most likely contrib- uting to corrosion of the stainless-steel sections along with the elevated acid gas content in the amine solvent. The most common pathway of chlorides ingression into amine units is dissolved in the produced water (brine) that enters with the inlet feed gas as water slugs, free water or aero- solized water. Ideally, liquid water should be removed prior to the amine unit using proper separation devices and gas-liquid coa- lescers. However, experience shows this is not the case in many facilities. Water slugs, pipeline foam, and vessel deficiencies can all cause produced water (brine) with high levels of salts and other dissolved contaminants to ingress the amine unit. In a future article, the possibility of using to an alternative amine solvent such as MDEA to mitigate cor - rosion in the process to acceptable levels will be explored. David B Engel is Managing Director at Nexo Solutions. He has been awarded 21 US Invention Patents and authored more than 100 papers and conferences. He holds a PhD in organic chemistry from Indiana University Bloomington. Email: david.engel@nexosolutions.com Scott Williams is a Process Engineer, responsible for engineer - ing, development, and onsite testing at Nexo Solutions based in The Woodlands, Texas. He holds a BS in chemical and biological engineering from University of Colorado at Boulder. Cody Ridge is a Chemical Engineer from Texas Tech and a lead process engineer at Nexo Solutions, responsible for engineering, development and sales. He has more than 10 published papers and has worked with more than 50 gas plants and refineries.

Reactor & Tower Internals, Custom Fabrication

Solutions at Your Service.

Woven Metal Products (WMP) offers complete vapor-liquid and liquid-liquid separation elements and non-pressure retaining housing solutions for all tower and vessel applications. Our ISO-9001-certified fabrication facility features the latest equipment for knitting, forming and punching.

Contact WMP today for Superior Separations & Filtration Products.

Wire Mesh Mist Eliminators Liquid Coalescer Devices Housing and Supports Chevron Vanes Inlet Devices

sales@wovenmetal.com wovenmetal.com

+1 281-331-4466 Alvin, Texas, USA

USA

Reliable Quality Partnership Attitude Attention to Detail

www.digitalrefining.com

100% USDA PRIME

FTC filtration keeps your cash cow flowing

FTC filters provide consistent effluent quality with the most predictable filtration of impurities as feedstocks change in renewable fuel production. Keep your cash cow flowing with reliable pre-treatment of your catalyst feed. Beef up your filtration system. And eliminate unplanned shutdowns.

713.849.0849 ftc-houston.com

Maximising value from refinery off-gases

Case studies examine reactor designs shaped by plant needs and gas composition, demonstrating how ROG purification offers compelling economics

Wolf Spaether, Holli Garret, Kristina Morgan and Felix Schulz Clariant

I n the past, refineries viewed off-gases from fluid catalytic cracking (FCC) units, coker units, and similar sources sim- ply as waste streams, burning them as fuel gas or releasing them through flaring. Today, these off-gases are recognised as valuable resources, containing a rich blend of hydrocar- bons, olefins (as much as 30 mol% in off-gas), diolefins, and hydrogen, alongside some undesirable impurities. Refinery integration with ethylene plants aims to max- imise olefin yields. Treating off-gases for removal of critical impurities for the purpose of recovering high-value compo- nents such as ethylene, propylene, paraffins, and hydrogen can be a major part of this strategy to significantly boost the plant’s economics while reducing the CO₂ footprint. However, off-gas compositions vary significantly, espe- cially when factoring in the removal of associated impuri- ties. Both catalytic and adsorptive treatments are essential yet challenging to implement. Currently, across the industry, the proprietary nickel-based

OleMax 100 catalyst series treats more than 1,000 met- ric tons per hour of predominantly refinery-sourced off- gases for nitric oxide, oxygen, acetylene, and heavy metals removal for recovery of hydrocarbon products. This results in more than 300 metric tons per hour of ethylene capac- ity gained and improved process safety within the down- stream cryogenic processing section. In addition to safely removing contaminants, treating and recovering the valu- able components from off-gases with adsorbents and cat- alysts provides added benefits of reducing carbon dioxide and other pollutants emissions that are typically created when used in the refinery fuel gas and flare systems. Against this backdrop, a focus on experience and results in designing new and revamped off-gas catalytic treatment systems is forthcoming. In some cases, these systems are also known as De-Oxo reactors. Through case studies, specialised catalytic reactor designs shaped by plant needs and gas composition are examined. The examples cover

PDH (Catofin)

Mon Pur (PolyMax ActiSorb)

Propylene Ethylene

Polypropylene ( Po l y Max)

ROG / ERU / PRU (OleMax)

O - gas

Polyethylene

C-C-C

Steam cracker light ends

Steam cracker furnaces

Mon Pur (PolyMax ActiSorb)

Naphtha

EO, glycols

recovery (OleMax)

Gasoline Jet fuel Diesel Fuel oil

C-C

Pygas (OleMax)

EDC / PVC (OxyMax)

ODH-E

Methane

Styrene (StyroMax)

Benzene

Aromatic complex

Polystyrene

Ethylbenzene

C-C

Ethane

Xylenes

PTA (H2Max)

PET

NG liquids

Toluene

ROG = Renery o -g as, EPU/PRU = Ethylene/propylene recovery unit, PDH = Propylene dehydrogenation, EO = Ethylene oxide, EDC = Ethylene dichloride, PVC = Poly vinyl chloride, PTA = Pu ri ed terephthalate acid, ODH-E = Oxidative dehydrogenation of ethane

Figure 1 Refinery off-gas treatment is an important link in the crude to chemical value chain where low-value stream contain high-value olefins

Gas 2025

www.digitalrefining.com

volumes and compositions. The primary sources are the crude oil distillation CDU (atmospheric and vacuum), coking units (delayed and fluid), and fluidised catalytic cracker (FCC) sections. Other sources may involve hydrocracking, hydro - treating, reforming, and gas processing units. Clariant has come across combined off-gas streams as well as separated fractions (saturated and unsaturated). Due to the diverse range of feedstocks, ROG compositions vary significantly. However, drawing from three decades of industry experience serving multiple units, Clariant can provide insight into the typical composition patterns it has observed. It needs to be noted that the analysis of ROG compositions is difficult given the many components and impurities down to ppb levels (see Table 1 ). In many cases, the refinery and its respective engineering, procurement, and construction (EPC) partners provide expected/simulated ROG compositions in the absence of real analytical data. Trace impurities are rarely analysed and require offline laboratory test equipment una - vailable on-site. Clariant has offered to receive real ROG feed samples for trace analysis and testing in its laboratories to support a streamlined design for its clients. With modernised global exploitation of various quality oil reserves and the increasing focus on circularity, the authors believe that the composition of ROG and other off-gases may change due to more complexity and, most importantly, higher levels of contamination such as heavy metals, alkali metals, sulphur, and nitrogen compounds. Why ROG purification? Historically, ROG served primarily as fuel gas for refinery operations, powering furnaces, boilers, and process heaters. However, evolving economic and environmental imperatives have transformed this practice from a practical solution to the efficient use of valuable resources. ROG streams contain valuable components, including eth - ylene, propylene, and hydrogen – key elements in modern crude oil-to-chemical (COTC) operations. While stream com - positions vary significantly across applications, all require the removal of common contaminants such as oxygen, nitric oxides, and acetylene. These compositional variations dic- tate specific catalyst, adsorbent, and process configurations for optimal recovery. Modern uses of ROG include further processing and/or treatment to separate and recover valuable components to

ROG compositions with multiple components and impurities

Component Hydrogen H₂ Oxygen O₂ Water H₂O

Averaged, mol%

Range, mol%

20.57

9.5-30.0 0.0-0.5 0.0-0.88

0.12 0.49

Nitric oxide NOx

<1 ppm

n.a.

Carbon monoxide CO Carbon dioxide CO₂

0.43 0.14

0.1-1.9 0.0-0.6 1.8-40.0 0.0-0.31 1.2-35.0 0.6-30.5 0.0-0.2 0.4-28.0 1.1-11.7 0.1-12.6 0.0-3.8 0.0-0.89

Methane CH₄ Acetylene C₂H₂ Ethylene C₂H₄ Ethane C₂H₆

21.73

0.09

21.63 14.28

MA/PD C₃H₄

0.04 8.98 3.00 3.37 0.86 0.13

Propylene C₃H₆

LPG

C₄ unsaturated

C₅

C₆+

Poison

Range, mol ppb

Phosphine PH₃

n.a. n.a. n.a.

1-1,000

Arsine ASH₃ Mercury Hg

0-350 0-100

Table 1

both adsorbent use for removing toxins (mercury, arsine, and phosphine) and catalytic solutions for eliminating nitric oxides, oxygen, and acetylene. Refinery off-gas positioning Figure 1 demonstrates the integration of refinery operations into the petrochemical value chain with the major down- stream uses of the most important chemical building blocks, ethylene and propylene. The refinery off-gas (ROG) purifi - cation section highlighted in green may be integrated with so-called ethylene recovery units (ERU) and/or propylene recovery units (PRU). In some cases, the purified ROG is sent to the separation section of an ethylene plant (steam cracker), including the cryogenic part, which is highly safety-relevant and requires stringent control of critical impurities. It can also be a standalone unit where typically large vol - ume off-gas streams are treated in dedicated ERU/PRUs to capture the valuable components. What is refinery offgas? ROG is derived from several processing sections in varying

Product recovery (Example)

Hydrocarbon feed or o-gas from renery or multiple sources

Quench & fractionation

Furnaces

Compression

Methane / fuel gas H

Sulphur removal & drying

Ethylene Ethane Propylene

OleMax 100 Series Ni-based catalysts

Cold recovery

Benets and considerations

NOx and O removal while also converting Ac , MA , PD using n ickel - based catalyst that is sulphur tolerant

Well suited for integrated operating complexes ( r efinery+steam cracker) Excellent performance with complex o-gas composition Improves plant economics

Propane C+

Figure 2 Category #1: Complete catalyst system to meet specifications using catalystic selective conversion of NOx, O₂, Ac, MA, PD

Gas 2025

www.digitalrefining.com

Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20 Page 21 Page 22 Page 23 Page 24 Page 25 Page 26 Page 27 Page 28 Page 29 Page 30 Page 31 Page 32 Page 33 Page 34 Page 35 Page 36 Page 37 Page 38 Page 39 Page 40 Page 41 Page 42 Page 43 Page 44