July 2017

Bonus Report: LNG

Monetize coal seam gas via LNG with optimized treating and liquefaction processes

Coal seam gas (CSG) has been established as a viable source of natural gas. The monetization of CSG resources through LNG projects experienced rapid expansion in recent years as a result of increasing LNG prices.

Mokhatab, S., Consultant; Towler, B. F., University of Queensland

Coal seam gas (CSG) has been established as a viable source of natural gas. The monetization of CSG resources through LNG projects experienced rapid expansion in recent years as a result of increasing LNG prices. However, CSG development and production have recently dropped due to a glut of LNG supplies.

LNG production plants fed with CSG require fewer gas processing steps compared with many conventional natural gas sources. The less complex processing demands can improve operations and reduce costs.

Here, a critical overview of commercially proven processes available for the treating and liquefaction of CSG in a baseload LNG plant is presented. A discussion is also offered on selection issues for the main technologies that affect LNG plant configuration.


CSG, also known as coalbed methane (CBM), is a methane-rich natural gas formed from the decay of organic matter in anaerobic (low-oxygen) environments and trapped in underground coal seams by water and ground pressure, where the methane content of a coal seam is a function of its depth and rank. In practice, the economic target depth to extract CSG is between 400 m and 800 m, because deeper wells are more expensive to drill and shallower coals do not contain sufficient adsorbed gas. However, the optimum depth for each CSG field depends on many factors in addition to the properties of the coal itself.1

Fig. 1. CSG production well diagram.
Fig. 1. CSG production well diagram.

CSG, which is an unconventional gas source, often requires highly invasive extraction methods. To extract the gas, a steel-encased hole is drilled into the coal seam. An electrical submersible pump is used to move the associated (formation) water through the production tubing, located in the center of the production casing. As the pressure within the coal seams declines due to the pumping of water from the coal, both gas and produced water move through the naturally occurring microfractures (i.e., the “cleats”) and rise to the surface (Fig. 1). Over time, water production decreases and gas production increases from each well (Fig. 2).

Following separation of the gas and associated water at the wellhead two-phase separator, the produced gas at near-atmospheric pressure is sent to a central field gas processing facility, where it is dehydrated (usually by triethylene glycol absorption) to pipeline specifications, and then compressed prior to entering the main natural gas trunkline. The produced groundwater from coal seams may be fresh (suitable for direct use), or it may be contaminated and require treatment to remove salts and other chemicals prior to use or discharge to an existing water course.

Fig. 2. Production profiles for a CSG well.
Fig. 2. Production profiles for a CSG well.

CSG production is characterized by low-pressure, low-production-rate wells spread over a large geographical area. Many wells are required to extract commercial quantities of CSG since the reserve ages and gas production declines after 5 yr–10 yr (Fig. 2). CSG production projects are characterized by high uncertainty in production rates from individual wells, as well as a wide range of operating conditions from field startup to end of life. The geographically distributed, capital-intensive nature of these projects places intense focus on the intelligent design and optimization of upstream facilities to maximize gas production while minimizing capital costs and environmental impacts.1

Depending on the characteristics of the coal seam, gas extraction may require further stimulation techniques, such as hydraulic fracturing, to increase gas flow. Hydraulic fracturing of low-permeability formations is a well-known and widely used technique in which the producing formation is fractured by hydraulically pressurized water containing a proppant (such as sand) and chemical additives. This method eases the passage for the gas to migrate through the seam and increase well productivity. Although hydraulic fracturing is controversial, instances of contamination of water supplies are extremely rare and mostly unproven.

Gas production from coal seams has several advantages (e.g., easy to drill and complete the wells at relatively low costs). However, it does pose some unique production and environmental challenges—such as aquifer depletion, water treatment and disposal and leaking methane to atmosphere—that impact large areas of the surface. These issues must be considered and resolved in any CSG production project.

CSG composition

The produced CSG is typically methane rich, sweet [very low in carbon dioxide (CO2) and hydrogen sulfide (H2S)] and lean (very little ethane, propane and heavier hydrocarbons) (Table 1). However, significant variations exist in the relative proportions of methane, CO2 and nitrogen (N2) between different locations, due to differences in the nature and method of CSG extraction. Commercially viable CSG resources generally have a methane content in excess of 95 mol%, which makes CSG a good feedstock for LNG production; however, it will not be able to self-source hydrocarbon refrigerant.

The coal has a different sorption affinity for different gases, leading to changes in the composition of the produced gas with time. Although the hydrocarbon content of produced CSG changes very little over the production lifetime of a well, the concentrations of CO2 and N2 change significantly with time. Since the coal has a higher affinity for CO2 than N2, the N2 content tends to decline over time, whereas the proportion of CO2 increases. The higher production of CO2 represents a potential corrosion issue on wellhead/production facilities, since water is present.

The CO2 will also have a major impact on project economics, since treatment of high-CO2 gas is capital-intensive. This means that understanding the coal seam gas composition and its variation with time is key to ensuring efficient downstream processing operation, resulting in the implementation of a fit-for-purpose, capital-efficient CSG-to-LNG project.

Developing CSG-to-LNG projects

While several countries have CSG resources, the world’s largest proven reserves are in Russia, Canada, China, the US and Australia.2 Other than the US and Australia, little activity has been undertaken in the area of CSG production, mainly due to a lack of immediate need, coupled with low gas prices.

Although much CSG is already produced and used for domestic consumption, it is only within the last decade that the scale of CSG production in Australia has increased substantially. The establishment of a major new LNG export industry based on CSG feedstock is delivering economic benefits to the country. For example, on Curtis Island, offshore Queensland, three major LNG projects (Queensland Curtis LNG, Gladstone LNG and Australia Pacific LNG) represent the first applications of converting CSG into LNG.

These plants (each of which have two medium-scale LNG trains) will collectively account for more than 25 metric MMtpy of LNG production capacity.3,4 Project reviewers4 predict that, by the time the LNG plants on Curtis Island are fully operational, gas demand in the state of Queensland will have increased from 1 Bcfd to approximately 5 Bcfd.

Two main reasons are immediately evident for selecting a medium-scale (1.5 metric MMtpy–4 metric MMtpy) LNG production facility. First, drilling and completing the several thousand CSG wells required to support an LNG plant at higher production rates is a capital-intensive and time-consuming task. Another factor is the presence of groundwater in the coal seams, where a considerable quantity of water must be pumped from each CSG well before gas production can begin. This may take several months to several years per well. Once gas production is underway, it is essential to avoid interruptions in gas extraction, as groundwater can rapidly resaturate the coal seams.

To maintain the CSG wells in operation instead of shutting them in, certain strategies are considered, including line packing, gas injection for buffer storage, feeding the gas into the domestic natural gas grid, and flaring. These strategies may be appropriate options if the LNG production plant must be shut down for a short period, such as a few days or weeks. This will encourage LNG plant designers to consider medium-scale LNG trains, allowing the plant to continue production when one of the trains is down for maintenance or an unexpected shutdown. However, the improved plant reliability and availability must be evaluated against any negative consequence to project costs due to increased equipment count.2

Gas processing. Several design considerations must be properly addressed at the conceptual design stage so that key decisions can be made for developing a CSG-to-LNG project. Some of these considerations are described in the following sections.

Gas pretreatment. The LNG production plant requires an energy-efficient pretreatment package to remove contaminants and deliver feed gas with the required specifications to the natural gas liquefaction unit. However, in a CSG-to-LNG project, a lean CSG feed gas received from the main trunkline permits the simplification of the pretreatment scheme.

The acid gas removal unit (AGRU) removes acidic components, such as H2S and CO2, from the feed gas stream to meet the sales gas H2S specification and to avoid CO2 freezing and blockages in the cryogenic exchanger. A feature of CSG is the virtual absence of H2S that simplifies the design of the AGRU, which can then be optimized for CO2 removal alone.

Usually, promoted methyldiethanolamine (MDEA) solvents are used to meet the tight CO2 specification for LNG plants (50 ppmv), where a feed gas with 10 mol%–12 mol% CO2 can be handled by a promoted MDEA process. For higher-CO2 gases, a physical solvent process can be used to reduce the CO2 content to 1 mol%–2 mol% before treatment by the promoted MDEA process. The physical solvent process uses pressure letdown for solvent regeneration that does not require heating, and it can significantly reduce environmental impact from burning fuel for amine regeneration.5 Note: The discharged acid gas stream can be vented directly to atmosphere or reinjected underground without further treatment. However, a provision may be included for a final incineration stage if some allowance for trace quantities of H2S must be embodied in the design. This means that the complexity and unreliability associated with a sulfur recovery unit (SRU) is avoided altogether.

Molecular sieves are used to dry the CSG leaving the AGRU to below 0.1 ppmv to avoid hydrate or ice formation in the downstream cryogenic system. Although oxygen (O2) is seldom present in CSG, the presence of O2 at low levels in the feed gas (and in the regeneration gas, by a number of routes) is recognized as an issue. In fact, during the heating stage of the molecular sieve regeneration cycle, the O2 reacts with hydrocarbons (methane) at temperatures above 200°C to form water and CO2, where the formation of water interferes with the complete regeneration of the molecular sieves.

In any case, the most efficient solution will dramatically decrease the heating temperature to 190°C–200°C. At these temperatures, the combustion reactions will be nearly eliminated. Of course, these lower regeneration gas temperatures can adversely affect the outlet dewpoint if a 4A molecular sieve is used. The use of a 3A type molecular sieve is recommended, since 3A sieves have a lower maximum allowable temperature for regeneration than 4A material. The 3A sieves can help achieve the required outlet water dewpoint, even at these lower regeneration gas temperatures. As an alternate, a closed-loop regeneration system can be used to avoid the use of O2-laden dried gas for regeneration. However, this would add complexity to the regeneration circuit.

Mercury (Hg), if present in the feed CSG, should be removed to avoid the risks of Hg attack on the brazed aluminum heat exchangers and equipment in the cryogenic section, as well as to prevent environmental and safety hazards. For this reason, the LNG plant is designed conservatively, requiring Hg removal to levels below 0.01 µg/Nm3. A few options are available for Hg removal utilizing nonregenerable Hg sorbents and regenerable Hg adsorbents.5 However, the use of a special class of molecular sieves developed to perform simultaneous removal of water and Hg yields good Hg removal performance, with a potentially longer service life. Note: The handling of Hg-laden regeneration gas is hazardous and requires a special safety procedure.

Since there are no heavy components to remove from the feed CSG, condensate stabilization or NGL fractionation facilities are not required in the LNG plant. However, because the hydrocarbon liquids are not available for onsite extraction from the feed gas, they will need to be imported to the site for use as refrigerant components in the cryogenic refrigeration circuits.

Gas liquefaction. Three main types of refrigeration cycle (cascade, mixed refrigerant and turboexpander) have been proposed for onshore LNG applications. The different liquefaction cycles are compared in Table 2. In general, expander cycles are favored on smaller peakshaving facilities, and single mixed-refrigerant (SMR) cycles are more attractive for midscale LNG plants. For baseload LNG plants, propane-precooled mixed refrigerant (PPMR), cascade or dual mixed-refrigerant (DMR) cycles are favored due to their higher efficiency.

In selecting the most appropriate liquefaction technology, both technical and economic considerations must be addressed. However, in developing any CSG-to-LNG project, the project team will place high emphasis on the availability and reliability of the first train of the LNG plant to minimize the risk of a serious interruption of LNG plant operations and a subsequent shutdown of the CSG production wells.

In this respect, one processa can improve the overall integration of CSG production and LNG plant operation. The process is designed around a “two-train-in-one” concept to ensure superior reliability and availability. It can help reduce both planned and unplanned downtime of the plant to keep gas flowing from the CSG wells.

The process utilizes highly reliable and efficient aeroderivative gas turbines as compressor drivers. The turbines provide higher availabilities (typically over 95%) and high plant turndown ratios of up to 10% over long-term operation.7 As the air temperature increases in warm climates, the efficiency of the gas turbines decreases, with a consequent reduction in plant production that results in greater specific plant cost. In this case, inlet air chilling is an attractive approach that can considerably augment gas turbine power, resulting in an improvement to thermal efficiency and a reduction in CO2 emissions. This concept has been applied at all three Curtis Island LNG projects.3

Note: Other well-proven liquefaction technologies take into account design strategies to promote high availability and reliability factors. Each technology can be competitive within a certain range of train sizes. The ultimate choice of which process to select will remain dependent on project-specific variables and can be achieved only after a detailed study of all options is conducted.5

Nitrogen removal. The presence of more than approximately 1 mol% N2 in LNG may lead to auto-stratification and rollover in storage tanks, which presents a significant safety concern. A higher percentage of N2 content in the feed gas also impacts the liquefaction process itself by reducing liquefaction efficiency. In addition, high-N2-content feed gas may require treatment of the boiloff gas (BOG) for use as the fuel gas for the gas turbine(s) in the LNG plant. The need exists for an efficient technique for the removal of N2 from LNG, even for relatively low N2 levels.

For feed CSG containing N2 levels of approximately 1 mol%–2 mol%, N2 can be removed in the customary end-flash section within an LNG production plant. When N2 is present in higher concentrations, the end-flash process can utilize an atmospheric stripper rather than a flash drum in the end-flash process. The stripper operates at close to atmospheric pressure.

To create a stripping gas, a methane-rich side stream from the upper section of the stripper can be used. The tray liquid is heated by the LNG from the main exchanger, and then partially vaporized prior to feeding to the bottom of the stripper as a stripping gas. This simple stripping process can offer an efficient N2 separation from methane for LNG feed with N2 contents up to 10 mol%. The end-flash process requires a low-pressure exchanger and a column, which are lower-cost items. However, a disadvantage is seen in the generation of a high-N2-content, low-Btu-content flash gas that requires rich gas for blending to meet the gas turbine manufacturer’s fuel gas specifications.

For higher-N2-content (10 mol%–40 mol%) feed gas, N2 should be removed upstream of the liquefaction plant. The N2 removal plant requires high-pressure exchangers, fractionation columns and refrigeration compressors, which are costly pieces of equipment.

Several N2-rejection methods exist, such as cryogenic separation, membranes and molecular sieve technology. However, membranes and molecular sieve technology are bulk removal processes suitable only for small-scale plants. The only viable large-scale rejection technology is cryogenic separation.5

Several cryogenic schemes are known to reject N2; the process selection depends on the N2 content of the feed gas and the supply conditions. If rejected N2 is vented, then the hydrocarbon content (predominantly methane) of the N2 vent stream must meet the environmental regulation, which is typically set at 0.5 mol%–1 mol%. Therefore, the selection of the N2 rejection unit (NRU) process must consider the power consumption, the optimum hydrocarbon recovery to minimize the environmental impact, the operability and potential safety hazards, while maximizing performance and flexibility.

When the N2 removal is performed within the liquefaction section, it avoids the NRU product compression system (with refrigeration provided by a liquefaction unit refrigeration system), as well as the losses associated with reheating and cooling feed gas for N2 rejection. However, in this scheme, a high level of integration with the liquefaction system adds to process complexity and operation risk, as neither the NRU nor the liquefaction system is conventional.5

Fig. 3. Typical scheme of LNG production with CSG feed.
Fig. 3. Typical scheme of LNG production with CSG feed.

LNG production scheme. Commercial process technologies can be integrated and configured for various LNG production schemes, with each offering unique benefits. Fig. 3 shows a typical LNG production scheme based on a feed CSG. In these integrated LNG production schemes, the main objective is to generate an optimized solution to provide great process flexibility and systems reliability, while providing significant energy and capital cost savings. Note that the optimum solution will vary from project to project, as each feed CSG is different.


CSG reserves have grown to such an extent that a number of CSG-to-LNG projects have been developed around the world, particularly in Australia. A key step in the development of an attractive project of this type is the selection of efficient gas pretreatment and liquefaction technologies to best meet the project objectives.

Several technology options can be integrated into the design of the LNG plant in CSG-to LNG projects. When determining the optimal integrated scheme, cost, reliability and operational flexibility must be considered. HP


a The process described here is ConocoPhillips’ Optimized Cascade process.

Literature cited

  1. Garlick, P., N. Amott, S. van Wagensveld and P. Andrews, “Coal bed methane—Unconventional gas becomes an optimized solution,” GPA Europe Spring Conference, Leiden, The Netherlands, May 14–16, 2014.
  2. Unsworth, N. J., “LNG from CSG—Challenges and opportunities,” 16th International Conference and Exhibition on Liquefied Natural Gas (LNG 16), Oran, Algeria, April 18–21, 2010.
  3. Cathcart, A., H. Patel and M. Harbeson, “Successfully delivering Curtis Island LNG projects,” 18th International Conference and Exhibition on Liquefied Natural Gas (LNG 18), Perth, Australia, April 12–15, 2016.
  4. Towler, B., M. Firouzi, J. Underschultz, A. Garnett, W. Rifkin, J. Esterle, S. Tyson, H. Schultz and K. Witt, “The coal seam gas developments in Queensland,” Journal of Natural Gas Science and Engineering, Vol. 31, 2016.
  5. Mokhatab, S., J. Y. Mak, J. Y. Valappil and D. A. Wood, Handbook of Liquefied Natural Gas, Gulf Professional Publishing, Burlington, Massachusetts, 2014.
  6. Finn, A. J., G. L. Johnson and T. R. Tomlinson, “LNG technology for offshore and midscale plants,” 79th Annual GPA Convention, Atlanta, Georgia, March 13–15, 2000.
  7. Eaton, A., R. Hernandez, A. Risley, P. Hunter, A. Avidan and J. Duty, “Lowering LNG unit costs through large and efficient LNG liquefaction trains—what is the optimal train size,” 2004 AIChE Spring Meeting, New Orleans, Louisiana, April 25–29, 2004.

The Authors

Related Articles

From the Archive