Module 4 CO2 compression and transportation to storage site
Original text: S. Wong, APEC Capacity Building in the APEC Region, Phase II Revised and updated byand ICF International
After capture, CO2 needs to be transported under pressure to a long-term geologic storage site. This module addresses compression and transportation systems for CO2.
By the end of this module you will:
- Understand CO2 compression technologies and the main operating issues pertinent to CO2 transport;
- Know the key considerations which must be taken to compress CO2 for transport;
- Be able to determine optimal CO2 pressure for transport;
- Appreciate the various operational issues association with CO2 compression;
- Understand the factors which influence costs for CO2 compression; and
- Be able to assess the various options for transporting CO2.
Gas compression is well developed in the natural gas industry around the globe and it uses mature technologies. CO2 compression uses the same equipment as natural gas compression, with some modifications to suit the properties of CO2. Avoiding corrosion and hydrate formation are the main additional operating issues when dealing with CO2. Compressors come in many different types (e.g. centrifugal, reciprocating and others), makes and sizes. Centrifugal compressors are usually the preferred type for large volume applications because of their ability to handle large flow rate (to hundred thousands of cubic feet per minute). In 2002, 490,000 hp of compression was installed in the USA, with a capital investment of over $ 635 million (True, W., 2003).
In 1998, more than 25 million tonnes of CO2 were captured, compressed, transported and injected in the Permian basin in the USA to recover nearly 150,000 barrels/day of oil through CO2(EOR) schemes (Stevens et al, 2000). The longest CO2 pipeline is the Cortez pipeline (808km) which has been delivering 20 million tones of naturally sourced CO2 per annum since 1972 (WRI, 2008). This illustrates that in some areas of the world there is more than adequate operating experience in compressing and handling CO2 in large-scale applications. In addition, there is significant experience in Western in acid gas (CO2 and H2S) compression and injection into geologic for disposal, and some pipelines in the US also carry CO2 from anthropogenic sources. As the oil reservoirs are maturing, for example, those in and Indonesia, the prospect of injecting CO2 to enhance the oil production or for storage can be technically and economically feasible. If these projects are to be implemented, large-scale handling of CO2 is required.
Compressing CO2 for transport
CO2 is compressed to make it more efficient to transport. The amount of compression needed for transport can be calculated through use of a phase diagram. Figure 4.1 shows the phase diagram of CO2. A phase diagram is a pressure-temperature relationship in graphic form, which shows the boundaries of the three phases – solid, liquid and gas phases. There are two important points to note on the phase diagram – the triple point and the critical point.
The triple point occurs at a pressure of 0.52 MPa and a temperature of -56°C. At this point, solid, liquid and gaseous phases of CO2 coexist together. Below this pressure and temperature, CO2 can only exist in either the gaseous or the solid phase. The critical point occurs at a pressure of 7.38 MPa and a temperature of 31.4°C. Above this critical pressure and at higher temperatures than –60°C, only one condition exists: that of the supercritical/dense phase. This kind of data can be found in engineering data handbook, for example, Gas Processors Suppliers Association (GPSA) Engineering Data book.
Figure 4.1: Phase diagram of carbon dioxide.
Figure 4.2 shows the density of CO2 as the pressures and temperatures vary. Above the critical pressure of 7.38 MPa and at temperatures lower than 20°C, CO2 would have a density between 800 to 1,200 kg/m3 (compare this to the density of water which is 1,000 kg/m3). A higher density is favourable when transporting liquid CO2, as it is easier to move a dense liquid than a gas. Therefore it is typical to compress CO2 to above 7.38 MPa for efficient transport. Typical pipeline temperatures are above 31oC, in order to maintain dense fluid characteristics of the CO2 stream.
Figure 4.2: Density diagram of carbon dioxide.
There is frictional loss as the CO2 flows through a pipeline. Typically the frictional loss can range from 4 to 50 kPa per km, depending on the pipe diameter, mass, CO2 flow rate and the pipe roughness factor. As a rule, the larger the pipeline diameter, the lower the frictional loss. Hence, in order to maintain the CO2 in thefor the whole pipeline, we would either maintain the inlet pressure to the pipeline at a high enough pressure to overcome all the losses while still above 7.38 MPa or install booster stations every 100 to 150 km to make up the pressure losses. Industry preference is to operate the pipeline at greater than 10.3 MPa at the inlet (i.e. the compressor discharge pressure) so that the CO2 would remain in the supercritical phase throughout the pipeline. It should be noted that when CO2 remains in the dense phase, we could revert to pumping rather than compression to achieve the higher pressure needed.
It should also be noted that maintaining the CO2 stream in a dense phase requires that the other incidental substances (impurities) in the stream be limited, as the phase diagram can be altered by these incidental substances, such as nitrogen, argon, water, etc. See below.
Getting to the CO2 pressure for transport
A number of stages of compression will be required before an optimal pressure is achieved for transport of the CO2. This is because, from engineering principle, it is impossible to raise the pressure of a gas such as CO2 ten to twenty fold in one step, as this would result in too high a temperature rise in the gas. Therefore, compression generally occurs in a number of steps or stages.
To determine the number of stages of compression, it is first necessary to select the acceptable compression ratio per stage. This ratio is generally in the order of 3 or 4. Table 4.1 shows the suction and discharge pressures and temperatures of a four-stage compressor with a compression ratio of 3. If needed, a fifth stage of compression can also be added.
As can be noted in Table 4.1, there is a considerable rise in the gas temperature during each stage. Aerial coolers are generally used to cool the process CO2 stream to the appropriate suction temperature between stages. There are also line and intercooler pressure losses at each stage. With a reciprocating compressor capable of a compression ratio of 4 per stage, a discharge pressure as high as 33 MPa can be achieved for a four-stage compressor.
Table 4.1: Suction and discharge pressures and temperatures as compiled through the computer model simulation.
Four to five stages of compression are required to obtain CO2 at optimal transport pressure. The compression ratio is used to determine this.
Typically the CO2 compression efficiency is about 80%. For example, the energy required for compressing CO2 to 14 MPa would be about 119 kWh per tonne of CO2.
Operational issues associated with CO2 compression
Since CO2 dissolves in water and forms carbonic acid, which is corrosive, strict control of the water content in the CO2 stream is essential for safe and efficient operation of the compressor.
In some plants, the required limitation on water content in the CO2 pipelines can be met in the water gas shift reactors and acid gas removal stages. In other plants, additional steps must be taken to remove water from the CO2 stream (European Commission, 2011). If the CO2 stream is kept well below the critical temperature (31oC) during the compression/pressurization process through the use of multiple compression stages with intercoolers, the liquid water and water vapour can be removed in dehydration units between the compressor stages (DNV, 2010). When the CO2 temperature rises above the critical temperature, the supercritical CO2 can absorb much more water than liquid CO2 at lower temperatures. If this occurs, the dehydration units (e.g., glycol dehydrators or molecular sieves) must be installed upstream of the compression stage.
In fact, a glycol dehydrator is often installed ahead of the CO2 compression stream to control water content to an acceptable level. Glycol dehydration is a standard unit operation in the oil and gas industry. For smaller scale operation, the dehydration unit may be eliminated if the temperatures of the gas stream at the intercooler stages can be controlled to drop off the water. This technique is widely practised in acid gas compression in Western Canada.
In addition, incidental substances, such as H2S or SOx in the presence of water can result in sulphuric acid, which can be quite bad for corrosion.
When dehydration is included, the metallurgy of the compressor piping can be relaxed. However, whether to switch back and forth between carbon steel and stainless steel or whether to make all piping around the compressor out of stainless steel depends on the cost difference. If the cost difference is small, it may be more practical to use all stainless steel in all the piping, coolers and suction scrubbers. Even though there may not be any water present in certain lengths of the piping, H2S (when present) reacts with carbon steel to form a thin film of iron sulphide on the surface of carbon steel. The iron sulphide may be dislodged at times and coat the inside surface of the stainless steel aerial coolers, thus decreasing the heat transfer efficiency. To avoid this potential heat exchanger problem, it may be advisable to use stainless steel throughout the compressor piping if H2S is present in the CO2 stream.
Sealing materials and gaskets
In addition, some petroleum based and synthetic lubricants can harden and become ineffective in the presence of CO2, so specific sealing materials and gaskets are typically specified in the USA for CO2 compressors and pipelines. CO2 cools dramatically during decompression so pressure and temperature must be controlled during routine maintenance (Gale et al., 2003).
Depending on the source of the flue gas, the CO2 stream recovered from it may contain trace concentrations such as H2S, SOx, NOx, O2, N2 and Ar. These impurities might have an impact on the physical state of the rich CO2 stream and hence the operation of the compressors, pipelines and storage tanks. The impact of impurities on CO2 transport is an ongoing topic of research.
A 2008 study from the University of Newcastle highlights the effect of impurities present from different sources (Seevam et al, 2008). Repressurisation distance depends on composition defined for the pipeline and the resulting thermodynamic properties (see Table 4.2).
Table 4.2: Properties of various capture streams (Seevam et al, 2008).
Table 4.3 show the pipeline quality specifications for existing US CO2 pipelines.
Table 4.3: US CO2 Pipeline Quality Specifications (source: INGAA Foundation, 2008)
A recent Guidance Document from the 7has provided additional information on the challenges with impurities in the CO2 stream.
Integration of capture, transport, and storage
Care must be taken to carefully integrate capture and storage specifications to get effective overall capture costs down. Furthermore poor selection of pipeline compositions can dramatically affect storage effectiveness through lower CO2 density – approx halved between 85-90%.
This demonstrates that if gas composition is not considered in transportation design considerable extra costs will result. Source to sink thinking is vital to CCS economics. While screening will be required for all installations it is highly likely higher concentrations will be indicated for all capture technologies. See Module 12 for further discussion on costs and source-sink economics.
CO2 can be transported by land via pipelines, motor carriers or railway, or by ocean via ships.
Three potential systems can be considered for land-based CO2 transport, namely: motor carriers, railway and pipeline.
Pipelining is currently the most economical method of transporting large quantities of CO2, and therefore the preferred option. There are currently some 6,200 km of CO2 pipelines in operation in the USA and Canada, transporting 30Mt per year of CO2 (IEA, 2009). These pipelines transport CO2 in the supercritical or dense phase.
Pipeline costs come from: material, labour, right-of-way access and miscellaneous. The average cost per mile shows few clear-cut trends related to either length or geologic area. In general, however, the cost per mile within a given diameter indicates that the longer the pipeline, the lower the unit cost (per mile) for construction. And, lines built near populated areas tend to have higher unit costs. Additionally, road, highway, river or channel crossings and marshy or rocky terrain each strongly affects the pipeline construction costs (True, 2003).
Liquefied CO2 can be transported in motor carriers such as tank trucks with trailers and stored in cryogenic vessels. The tanks have an inner vessels or "liquid container" which is surrounded and supported by an outer vessel or "vacuum jacket". The space between the two is filled with a natural material that provides insulation. The delivery system includes piping which carries gas from the vessel through the vacuum jacket to the outside, controlled by gauges and valves mounted outside the tank. These vessels are available in various sizes ranging from 2 to 30 tonnes, to suit customers' requirements. The conditions of the liquid CO2 is typically at 1.7 MPa, -30°C. Currently these are used to transport CO2 for the food and beverage industries, but the volumes are very small compared with what will be required for CCS.
This kind of vessel offers flexibility, adaptability and reliability. This is the most common form of bulk CO2 transport for retail purposes. Figures 4.4 and 4.5 show a fleet of the CO2 tanker trucks and refilling a truck with CO2.
Figure 4.4: A fleet of CO2 tanker trucks for oil field applications in China.
Figure 4.5: Refilling a CO2 tanker truck.
The railway system has a large carrying capacity that enables it to handle large volumes of bulk commodities over long distances. CO2 can be transported in specially developed tank cars that are approved to transport liquid CO2 at a pressure of 2.6 MPa. At this pressure, the net weight of liquid CO2 that a single tank car can load is about 60 tonnes. CO2 has been shipped by rail in two or three tank cars at a time. The tank cars can be left at the siding and serve as a storage tank until the next shipment arrives. However, there is no large scale CO2 transport by railway at this point. Rail transport will only become a competitive transport option if the logistics can fit the volumes in the existing railway system. However, loading and unloading infrastructure and temporary CO2 storage would also have to be included in calculating the cost.
Ships can be used for long distance transport of CO2 across oceans. Smaller dedicated CO2 ships are in operation today. The size of these ships is up to 1,500 m3, and the transport pressure is about 1.4 to 2 MPa. These ships are not suitable for large-scale ship-based transport of CO2 because at these pressures, the ship must be constructed as pressure vessel, which will make costs very high. Lower pressure is required for enlarged storage and ship tanks. However, this should not be a major problem, as tankers are currently used for shipping liquefied petroleum gas (LPG) and tankers similar to these could be used for CO2. Figure 4.6 shows a LPG tanker.
Figure 4.6: A LPG tanker – CO2 could be transported in a similar way. Courtesy of Mitsubishi Heavy Industries.
The Weyburn pipeline: case study of a compression and transport system Drawn from the work of Hattenbach et al, 1999
The Weyburn Pipeline is a 320 km CO2 pipeline, which stretches from the Great Plains Synfuels Plant (GPSP) near Beulah, ND to EnCana's oil field in Weyburn, Saskatchewan, Canada (see Figure 4.7).
Figure 4.7: Location of the EnCana Weyburn CO2 Pipeline (in white).
The pipeline is sized to handle the entire waste gas output of the Rectisol unit from GPSP (about 5 million tonnes of CO2 per year). Phase 1 of the pipeline consists of a 356 mm (14 inch) diameter section from the plant to near Tiogas, ND followed by a 305 mm (12 inch) diameter section from Tioga north to Weyburn field. Initial pressure of the CO2 leaving the plant is 17 MPa (2,500 psig) and the delivery pressure at Weyburn is 14.8 MPa (2,175 psig). Phase 1 is designed to deliver approximately 5,000 tonnes/day of CO2 to the Weyburn oil fields. As the pipeline diameter is sized much bigger than the initial delivery volume, the frictional loss is very low at 7 kPa per km. For this pipeline, there is no booster station for the entire 320 km pipeline.
For compression, a dual compressor train (Figure 4.8) was designed to handle the initial throughput of the 5,000 tonnes per day for the Weyburn EOR project. Each compressor train (3 stage compressor) has an initial capacity to handle more than 2,500 tonnes/day of CO2. A CO2 pump is used to boost the pressure to 17 MPa for pipeline delivery. More capacity can be added later by either adding the additional compressor trains or by installing booster compressors on one or more existing trains. Adding pumping stations on the pipeline would allow increased flow to be achieved in the pipeline itself.
The pipeline has been in operation since year 2000. The IEA launched an international research program, the IEA Weyburn CO2 Monitoring and Storage Project in 2001. This program considers how best to combine oil recovery and long term storage. In 2005, the Midale field nearby (operated by Apache Canada) became part of the project known as the IEA GHG Weyburn-Midale CO2 Monitoring and Storage Project and is in its final phase. EnCana is injecting 7000 tonnes/day and Apache is injecting 1800 tonnes/day (PTRC website).
Gas compression is a well-developed industry in North America, using matured technologies. Basically, CO2 compression uses the same equipment as natural gas compression, with some modifications to suit the properties of CO2. Avoiding corrosion and hydrate formation are the main additional operating issues involved with CO2.
The water content in the CO2 stream must be strictly controlled to prevent corrosion. A glycol dehydrator can be used for this purpose. To avoid potential heat exchanger problems, it is advisable to use stainless steel throughout the compressor piping if H2S is present in the CO2 stream.
Special sealing materials and gaskets are recommended in order to avoid hardening of some petroleum based and synthetic lubricants in compressors and pipelines. Impurities in the rich CO2 streams may impact on the compressor and pipeline operations. Their impact is currently being researched.
Pipelining is the most economical method to transport large quantities of CO2. Rail transport would be competitive if the logistics can fit the volumes in the existing railway system. However, loading and unloading infrastructure and temporary CO2 storage would have to be included and would be costly. Tanker truck transport would not be viable to transport large quantities of CO2, because of cost and volume considerations.
It is more efficient and economical to ship CO2 in the supercritical/dense phase. For pipeline transport, that means keeping the pressure at above 7.38 MPa.
In North America, we have adequate experience in compressing, pipelining and handling CO2 in a large scale. This experience can be easily transferred to APEC economies, should such a need arise. However each project should be considered in a full source to sink way.
Gale, J. and Davison J., Transmission of CO2 – Safety and economic considerations, Proceedings of the 6th International Conference onControl Technologies, J. Gale and Y. Kaya, editors, Elsevier Science, New York, 1: p517-522, 2003.
Gas Processors Suppliers Association (GPSA), Engineering Data Book, Volumes 1 and 2, revised tenth edition, published by GPSA, Tulsa, Oklahoma, 1994.
Hattenbach, R., Wilson, M. and Brown, K., Capture of carbon dioxide from coal combustion and utilization for enhanced oil recovery, in Proceedings of the 4th International Conference on Greenhouse Gas Control Technologies, P. Reimer, B. Eliasson and A. Wokaun, editors, Elsevier Science, New York, p217-221, 1999.
Heddle, G., Herzog H. and Klett, M., The economics of CO2 storage, Laboratory of Energy and the Environment, http://lfee.mit.edu/publications/ Publication No. LFEE 2003-003 RP, 2003.Institute of Technology, Cambridge, MA, USA 111p.
IEA Technology Roadmap - Carbon capture and storage. October, 2009. Available from www.iea.org
Odenberger, M. and Svensson R., Transportation systems for CO2 – Application to carbon sequestration, Chalmers University of Technology, Goteborg, Sweden, 48p. http://www.entek.chalmers.se/~klon/msc, 2003.
Seevam, P., Race, J., Downie, M. and Hopkins, P., Transporting the Next Generation of CO2 for Carbon, Capture and Storage: The Impact of Impurities on Surpercritical CO2 Pipelines, Proceedings of the IPC2008 7th International Pipeline Conference (Paper IPC2008-64063), 2008.
Stevens, S. and Gale J., Geologic CO2 sequestration, Oil and Gas Journal, p40-44, May 15, 2000.
Torp, T.A. and Brown, K.R., CO2 Underground storage costs as experienced at Sleipner and Weyburn, presented at the 7th International Conference on Greenhouse Gas Control Technologies, Vancouver, British Columbia, Canada, September 5-9, 2004.
True, W., US pipeline companies solidly profitable in 2002, scale back construction plans, Oil and Gas Journal, p 60-90, September 8, 2003.
World Resources Institute (WRI). CCS Guidelines: Guidelines for Carbon Dioxide Capture, Transport and Storage. Washington, DC: WRI, 2008.
World Resources Institute: www.wri.org
Weyburn-Midale CO2 Project: www.ptrc.ca/weyburn_overview.php
Current research at the University of Newcastle: www.ncl.ac.uk/marine/research/project/2000
The Cortez pipeline: www.kindermorgan.com/business/co2/transport_cortez.cfm