Carbon dioxide-tolerant nitrogen rejection reduces costs

A. Farooq, A. Finn, A. Hosainy and G. Johnson, Costain Natural Resources, Manchester, UK

For commercial-scale nitrogen (N₂) removal, cryogenic distillation is the only cost-effective technology. Cryogenic N₂ rejection produces a high-purity N₂ waste stream, which can be vented to the atmosphere and which achieves very high hydrocarbon recovery with low power consumption.

Cryogenic N₂-rejection processes that employ a pre-separation distillation unit can handle up to 60 times as much feed gas CO₂ as processes that do not use pre-separation. A new process, achieved through the modification of an established cryogenic N₂-rejection process, allows up to approximately 2 mol% CO₂ content in the feed gas to the cryogenic N₂-separation process. This allowance reduces capital and operating costs by eliminating or reducing the upstream removal of CO₂. It also eliminates the risks associated with the handling and storage of chemicals typically required for CO₂ removal.

N₂ is inert and does not burn, so natural gas containing excessive N₂ (typically more than 4 mol%–5 mol%) must be processed to remove N₂ to meet sales gas minimum heating value specifications. Gas fields with high N₂ levels exist in Siberia, Eastern Europe, the Midwest US, The Netherlands, the Irish Sea, the North Sea, Pakistan and Southeast Asia. In all of these locations, N₂ removal (rejection) is potentially important.¹

N₂ removal also reduces gas transportation costs. If natural gas is to be used as petrochemical plant feedstock, then the N₂ content must be even lower—potentially less than 1 mol%. The lower N₂ content is necessary because N₂ is very difficult to separate from carbon monoxide (CO) in synthesis gas formed by steam methane reforming. If not removed, N₂ can cause serious bottlenecks in petrochemical production.

N₂-removal technologies. The technologies available for N₂ removal from natural gas include pressure swing adsorption (PSA) and rubbery membranes. However, for commercial-scale N₂ removal (at flowrates greater than 20 MMscfd), cryogenic distillation is the only cost-effective technology.^{1, 2} Rubbery membranes are limited to much lower flowrates and are only economical for fuel gas conditioning. A crucial advantage of cryogenic N₂ rejection is that it produces a high-purity N₂ waste stream, which can be safely vented to atmosphere. 

Cryogenic processing also achieves greater than 99% hydrocarbon recovery, to maximize revenue. It has relatively low power consumption and small sales gas compressor(s). Alternative process technologies produce methane (CH₄)-rich N₂ waste streams that present disposal challenges and bigger compression requirements.

Cryogenic N₂ rejection units (NRUs) operate at temperatures as low as –190°C and require upstream removal of feed gas components that could freeze and cause equipment blockage. Water is easily removed by molecular sieves, and heavier hydrocarbons are removed by partial condensation or proven dewpointing techniques, such as silica gel adsorption.

CO₂ must also be considered, to prevent it from freezing. The CO₂ content of natural gas varies widely, but CO₂ is commonly present at levels up to several percent and above. These levels are too high for an NRU, as CO₂ solidifies at NRU operating temperatures. The amount of CO₂ tolerable in the NRU feed is limited by its solubility in methane at the low NRU temperatures.³ Therefore, CO₂ removal is typically required upstream of the NRU.

Depending on the NRU process configuration, the practical CO₂ limit in the feed gas may be as low as 50 ppmv, akin to that of an LNG plant. However, this is only true for very high-N₂-content gas, so considering CO₂ removal requirements in the same way as for LNG is erroneous. This is because, when the feed gas N₂ content is less than 30 mol%, a preseparation column is commonly used to remove methane at elevated pressure (and temperature). This significantly reduces power consumption, compressor size, capital cost and overall processing cost. The rectification section of the preseparation column removes CO₂ from the overheads, so much less CO₂ enters the colder downstream section of the NRU, and the plant feed gas can potentially contain up to 0.3 mol% CO₂ without creating any freezing concerns.

An NRU design can be made more CO₂-tolerant (up to the 0.3 mol% figure), and this can be beneficial to acid gas removal cost and/or provide a “buffer” to plant upsets. However, the issue remains that, while N₂ removal by cryogenic processing is clearly optimal, a cost burden normally exists for CO₂ removal. PSA technology can remove CO₂ as well as N₂, but, as noted previously, this potential advantage is outweighed by excessive compression costs at flowrates over 20 MMscfd. Therefore, efficient removal of N₂ from natural gas almost inevitably requires adjunct CO₂ removal, even though the sales gas CO₂ specification may be already met.

Several well-established processes for CO₂ removal exist, but all incur significant cost and may impact NRU reliability and operation. Therefore, an incentive exists to avoid CO₂ removal upstream of cryogenic NRUs by making the NRU design more tolerant to CO₂.

Conventional cryogenic NRU process. If the feed gas N₂ content is higher than around 30 mol%, as is typically found in Eastern Europe, then the N₂ removal is conventionally carried out in a cryogenic double-column process for excellent process integration, high thermodynamic efficiency and minimal power consumption for methane compression.⁴ The feed gas is subjected to very low temperatures, which limit the feed gas CO₂ concentration to 50 ppmv–100 ppmv; otherwise, freezing would occur.

Natural gas that requires N₂ removal normally has a much lower N₂ level than 30 mol%. In these applications, a preseparation distillation column is commonly used.^{5, 6} These columns operate at relatively high pressure and temperature, approximately 30 bar–35 bar and –115°C as the coldest temperature.

As its name suggests, the preseparation distillation column removes a portion of the feed gas CH₄ before the downstream N₂-CH₄ separation (to produce low-CH₄-content N₂). The downstream separation processes only a fraction of the feed gas, which greatly improves process efficiency and reduces power consumption and cost. The methane from the preseparation column is at relatively high pressure, so compression to sales gas pressure consumes less power (and makes the overall sales gas compression system simpler, in most cases). Fig. 1 shows a typical NRU flowsheet incorporating a preseparation column.

Fig. 1. Typical NRU with preseparation column.

NRUs are autothermal and do not require external refrigeration, as the required refrigeration is provided by one or more evaporating methane streams. Optimizing the overall separation system, the provision of refrigeration and the sales gas compressor system together greatly increases thermodynamic efficiency and reduces compression costs. Therefore, the processing costs of N₂ rejection have fallen significantly in recent years.

In Fig. 1, the feed gas is partially condensed by returning product streams, and it enters the preseparation column. This column has an overhead condenser to provide reflux to limit methane going downstream. It also limits the CO₂ content in the gas flowing to the colder downstream distillation.

The bottom product from the preseparation column contains typically 1 mol%–4 mol% N₂ and essentially all of the CO₂ in the feed gas, due to CO₂ being less volatile than methane. Evaporation of this stream provides refrigeration for condensing the feed gas and preseparation column reflux. The product from the downstream N₂-rejection column system is evaporated at lower pressure, and it provides refrigeration in the colder section of the plant. Once evaporated, the bottom products from both the preseparation and N₂-rejection section can be combined and compressed to sales gas pressure.

Since the preseparation column overhead reflux removes most of the CO₂ from the feed stream in the preseparation column bottoms (at relatively high pressure and temperature), increased feed gas CO₂ content to the NRU can be tolerated—up to around 0.3 mol%—when compared to a process without preseparation. The feed gas CO₂ limit is influenced by the N₂ content of the feed gas and the amount of methane removed by the preseparation column.

The limitation of 0.3 mol% is in place because the bottom stream from the preseparation column (dashed stream in Fig. 1) is let down in pressure to condense the feed gas and preseparation column overheads. If the CO₂ content in the feed gas were higher than approximately 0.3 mol%, then CO₂ would solidify on pressure let-down due to the temperature reduction from the Joule-Thomson effect. Therefore, if this pressure let-down could be avoided, then a higher CO₂ level—of up to 2 mol% CO₂—would be feasible; i.e., within solubility limits.

This increase in CO₂ tolerance could actually avoid the need for upstream CO₂ removal and provide significant cost savings. An alternative source of refrigeration would be required; however, if it can be provided relatively easily, using established principles and process equipment, then significant cost savings are feasible without increased technical risk.

CO₂-tolerant process. A modified NRU process has been developed that avoids the need for pressure let-down of the bottoms stream from the preseparation column by using an alternative source of refrigeration. A CO₂-free methane stream is sourced from downstream of the preseparation column and employed in a refrigeration cycle.⁷

Similar to the NRU process described above, the low-CO₂-content bottoms product from downstream distillation is evaporated to provide refrigeration. However, instead of combining it with the preseparation column bottoms to flow to compression, the low-CO₂-content stream is compressed in a separate stage in the compressor system.

A part of the low-CO₂-content stream from downstream of compression is then recycled, condensed, let down in pressure and evaporated at one or more pressure levels to provide refrigeration to condense feed gas and for the preseparation column condenser (see dashed line in Fig. 2). The evaporated stream is then returned to the compressor system. Although two different compressor systems are required, they may actually share a common driver to reduce costs.

Fig. 2. CO2-tolerant NRU.

A case study is presented for a natural gas stream containing 20 mol% N₂ and 1 mol% CO₂.⁷ A feed gas flow of 200 MMscfd at 45 bar and 45°C is assumed. Key process parameters are outlined in Fig. 2. The predicted CO₂ solubility is presented for the key process streams, where CO₂ freezing could potentially occur, in Fig. 3 (preseparation column, stage by stage) and in Fig. 4 (evaporating preseparation column bottoms stream).

Fig. 3. CO2 concentration vs. predicted solubility limit for preseparation column.

As can be seen in Figs. 3 and 4, ample operating margin exists between the predicted CO₂ concentration and the solubility limits, so no CO₂ freezing would occur for a feed with a CO₂ content of 1 mol%. Fig. 4 shows the predicted solubility and CO₂ concentration of the bottom product from the preseparation column. As discussed, let-down in pressure of this stream is the limiting factor for CO₂ freezing and, as such, limits the amount of CO₂ that can be present in the feed gas. The new process avoids such low pressure, and so the amount of CO₂ in the feed can be greater.

Fig. 4. CO2 concentration vs. predicted solubility limit in evaporating preseparation bottom product stream.

Fig. 4 also shows a predicted limit of around 5 mol% CO₂ in the bottom product from the preseparation column. Therefore, the CO₂ content in the feed stream can be higher than shown in the case study (up to approximately 2 mol%) before freezing is predicted. Accurate prediction of CO₂ solubility is ensured by using a fully rigorous solid-vapor-liquid fugacity methodology that has been benchmarked against leading CO₂ freezing prediction models.⁸

The net power consumption of the new process is the same as the conventional process for a feed gas containing 0.3 mol% CO₂. This is because the new process allows the CO₂-rich product stream from the preseparation column to be evaporated at elevated pressure, which reduces compression power and offsets the refrigeration cycle compression power. Also, the refrigerant stream is nearly pure methane, whereas the preseparation column bottom product can contain significant C₂+ content, affecting evaporating temperature and requiring lower evaporating pressure.

The key benefit of the new process is that, for natural gas streams containing less than 2 mol% CO₂, an upstream acid gas removal unit (AGRU) can be avoided. Eliminating CO₂ removal can result in significant cost savings, increased plant reliability, improved operability, reduced operator and maintenance burden, and improved safety, as there is no need to store and handle CO₂-removal chemicals. Not only is the AGRU cost avoided, but, since feed gas is normally warmed in an AGRU, the dehydration unit is smaller, resulting in additional capital cost savings. The new process introduces no additional risks to operability or performance.

With regard to capital costs, the estimated split of capital costs for an NRU system (based on gas turbine drive) is typically:

• AGRU: 20%
• Dehydration: 13%
• NRU: 22%
• Compression: 45%.

If the new process avoids the need for an AGRU, then, given the other savings on dehydration and allowing for the additional compression for the refrigeration loop, the typical overall capital cost reduction will be approximately 10%–15%. Additional savings on utility heat, operating and maintenance costs could reduce processing costs by 15%–20%.

Established cryogenic NRU technology. Cryogenic NRU technology has been successfully implemented in major gas processing facilities, characterized by optimized distillation designs and energy integration by using multi-stream aluminium plate-fin heat exchangers. These plants are said to exhibit high energy efficiency, low compression power, very low hydrocarbon losses and excellent turndown characteristics. They provide a high rate of return and low processing costs, with minimal environmental impact. The new process is based on these well-proven, high-performing and reliable NRUs.

Fig. 5. Connah’s Quay Gas treatment plant, with preseparation column.

The Connah’s Quay gas treatment plant (E.ON) in Wales, UK handles up to 200 MMscfd of natural gas containing 10 mol% N₂ and 0.3 mol% CO₂ (Fig. 5).⁹ A similar facility, for Eni Pakistan’s gas concession in the Bhit Mountains, treats 270 MMscfd of gas with 18 mol% N₂ content in two trains (Fig. 6).¹⁰

Fig. 6. Eni Pakistan gas processing facility.

The NRU supplied to PEMEX in Villahermosa, Mexico has two trains, each with a capacity of 300 MMscfd. The plant is designed to operate with increasing feed gas N₂ over time as N₂ breaks through in the associated gas (Fig. 7) from enhanced oil recovery.

Fig. 7. Cuidad PEMEX gas plant in Tabasco, Mexico.

All of these NRUs are proven in meeting product specifications reliably, safely and efficiently.¹¹ In all cases, the NRU facilities were optimized to give a conventional, efficient and reliable sales gas compression configuration.

Takeaway. Table 1 summarizes the approximate maximum allowable CO₂ in NRU feed gas for an NRU without preseparation, with preseparation and with the new process.

The new process introduces a relatively small modification by using an alternative source of refrigeration. This improves the CO₂ tolerance of a well-proven cryogenic NRU with no loss in operability, flexibility or hydrocarbon recovery, and with no increase in overall power consumption.

The new process can eliminate an upstream AGRU, which results in the following key benefits:

• Significant capital and operating cost savings
• Increased plant reliability and operability
• Reduction in operator and maintenance burden
• Avoid warming of feed gas by AGRU, resulting in smaller dehydration unit and less NRU heat exchange area.

The new CO₂-tolerant process allows CO₂ concentrations of up to approximately 2 mol% in the feed gas—a significant increase compared to the conventional approach. GP

ACKNOWLEDGMENTS

The authors would like to thank Terry Tomlinson for his assistance in preparing this paper.

LITERATURE CITED

¹Finn, A. J., “Rejection strategies,” Hydrocarbon Engineering, October 2007.

²Johnson, G. L. and T. Eastwood, “Nitrogen rejection from natural gas,” GPA Europe Spring Conference, Copenhagen, Denmark, May 27, 2011.

³“Section 16: Hydrocarbon recovery,” GPSA Engineering Data Book, 13th Ed., 2012.

⁴Cholast, K. and A. Kociemba, “Technology assessment for two-phase LNG expanders operating for ten years in gas liquefaction process,” 30th GPAE Annual Conference, Edinburgh, Scotland, September 18–20, 2013.

⁵Trautmann, S. R. and K. H. Janzen, “Innovative NRU design at Pioneer Natural Resources’ Fain gas plant,” 79th GPA Annual Convention, Atlanta, Georgia, March 13–15, 2000.

⁶O’Brien, J. V. and J. J. Maloney, “Continuous improvement in nitrogen rejection unit design,” Hydrocarbon Engineering, September 1997.

⁷Johnson, G. L. and A. J. Finn, “Process and apparatus for separation of hydrocarbons and nitrogen,” International Patent No. WO 2014/167283 A2, 2014.

⁸Eggeman, T. and S. Chafin, “Beware the pitfalls of CO₂ freezing prediction,” Chemical Engineering Progress, March 2005.

⁹Healy, M. J., A. J. Finn and L. Halford, “UK nitrogen removal plant starts up,” Oil & Gas Journal, February 1999.

¹⁰Millward, R. J., A. J. Finn and A. J. Kennett, “Pakistan nitrogen removal plant increases gas quality,” GPA Europe Annual Conference, Warsaw, Poland, September 21–23, 2005.

¹¹Wilkinson, D. and G. L. Johnson, “An Abu Dhabi case study,” Hydrocarbon Engineering, February 2012.

Adil Farooq is a process engineer at Costain in Manchester, UK, with responsibility for hydrocarbon processing projects, including conceptual design, Pre-FEED, FEED and detailed design. His three years with Costain have been spent on cryogenic gas processing projects. He has a bachelor’s degree in chemical engineering and a master’s degree in advanced chemical engineering, both from the University of Manchester, UK. Mr. Farooq is an associate member of the IChemE in the UK.

 

Adrian Finn is process technology manager at Costain in Manchester, UK, with responsibility for process technology development and commercialization, proposal management and supervision of feasibility studies, pre-FEEDs and FEEDs. He has spent 32 years with Costain, mainly on cryogenic gas processing projects. He has authored nearly 50 technical papers and holds nearly 20 granted patents in gas processing. He earned a bachelor’s degree in chemical engineering and fuel technology from Sheffield University, UK and a master’s degree from the University of Leeds, UK. Mr. Finn is a fellow of the Institution of Chemical Engineers (IChemE), a chartered engineer in the UK, a member of both the management and program committees of GPA Europe.

 

Ahmad Hosainy is a graduate process engineer at Costain in Manchester, UK. He has worked on various technical studies, pre-FEEDs and detailed design in hydrocarbon processing involving a range of gas processing technologies. Mr. Hosainy holds an MEng degree from the University of Manchester, UK. He is also an associate member of the IChemE.

 

 

Grant Johnson is front end solutions manager, responsible for activities in oil and gas, at Costain in Manchester, UK. He joined Costain in 1997, with a master’s degree in chemical engineering from the University of Cambridge, UK, and has since worked on a number of large gas processing facilities. He holds patents in the field of natural gas processing, covering nitrogen removal, liquefaction and liquids recovery. Mr. Johnson is a member of the IChemE, a chartered engineer in the UK.