C. Gould, Fluor, Aliso Viejo, California; and R. JOYNER and A. DESHMUKH, Fluor, London, UK
As the world continues along its energy transition trajectory, demand for blue hydrogen (H2) and reduced carbon emissions will make high-purity carbon dioxide (CO2) streams available in many process plants. This article investigates how these streams could be utilized to add value to projects.
With energy production continuing to be decarbonized and renewable energy sources (such as solar and wind) becoming significant sources of power, the ability of power generation networks to respond to fluctuations in power demand may become limited. In the UK, swings in power generation are provided predominantly using the combustion of natural gas in gas turbines. To reduce reliance on hydrocarbon energy sources, energy storage solutions will be essential to balance supply with demand.
Part 2 of this article considers the potential to utilize the storage of liquid or supercritical CO2 as a means of energy storage, and includes a concept for an integrated carbon capture, refrigeration and CO2 injection process.
The need for energy storage. Conventional coal or nuclear-based power generation can provide consistent baseline power supply, and systems such as gas turbines or hydroelectric plants are well suited to provide swings in power output to meet demand as it changes throughout the day. Renewable energy sources, such as solar and wind, can be unpredictable and inconsistent, which results in peaks and troughs in the power supply. While energy supplies from wind and solar can be reduced when excess power is available, there is a very limited ability to increase supply when demand is high.
FIG. 8 shows power generation in the UK in 2022. Baseline power generation is provided predominantly by nuclear power, with a reasonably consistent generation of 4 GW–6 GW. Wind power is far more erratic, with generation ranging from 3 GW to more than 15 GW. It is apparent that wind power is mirrored by natural gas power generation, showing how natural gas is filling in the peaks and troughs in wind-supplied power generation—when the availability of wind power is high, natural gas power generation is low, and vice versa.
FIG. 9 shows the daily average power generation for the UK in 2022. Power demand is low throughout the evenings before increasing at around 6 a.m., and then it remains reasonably consistent until there is another peak in the evenings. It can be clearly seen how natural gas is being used to provide this swing in power demand. The amount of power generated from all other sources (other than solar) has no correlation to the time of day and has very limited ability to increase the rate of generation in periods of high demand; therefore, natural gas generation is ramped up and down to meet changes in demand. FIG. 10 shows how the wholesale price14 of electricity is strongly correlated with power demand—another key factor that will be considered later in this article.
The UK is particularly reliant on natural gas to perform this role, as the country’s landscape does not lend itself to hydroelectric power generation—the only other large-scale power source that can easily provide swings in generation capacity. Additionally, the country is heavily reliant on wind power with plans to increase this further in the future. The UK government has pledged to increase offshore wind capacity to 40 GW by 2030, enough to provide the entire power demand of the country.15 It is important to understand that this is the installed capacity, but this can only be utilized when sufficient wind power is actually available.
Energy storage systems. The concept of energy storage is not new. The most common source of stored energy is electric batteries, which store electrical energy in electrochemical cells. Battery technology has advanced rapidly in recent years, driven largely by the demand for electric vehicles. While battery technology for large-scale energy storage shows potential, many challenges must be resolved. One key challenge is that battery life decreases over time. Tesla guarantees that the lithium-ion batteries used in its vehicles will retain 70% of their original capacity after 8 yr of use.16 While this is good performance for an electric battery, process plants are regularly designed with a life of 30 yr or more without a significant loss of capacity. There are also environmental concerns around the mining of metals used in batteries, such as lithium, cadmium and nickel. By comparison, CO2 process plants can be constructed from carbon and stainless steel.
Another common means of energy storage utilizes hydroelectric dams in which excess power is used to pump water up to a higher elevation where it is held in a reservoir, storing gravitational potential energy. This energy can then be released by allowing the water to flow through hydraulic turbines under gravity, generating electricity. This form of energy storage is highly limited by the landscape.
The purpose of this article is not just to provide an in-depth discussion of alternative forms of energy storage, but also to highlight some drawbacks and challenges of existing energy storage systems. The following sections will discuss the viability of CO2 energy storage, and how it may avoid some of these issues.
Liquid CO2 for energy storage. The use of liquefied gases as a means of energy storage is an established concept. Liquid air energy storage (LAES) technology uses stored thermal energy in the form of liquid air, which can then be released by vaporizing the air against a low-grade heat medium and passing it through a turbine.17 CO2 has a significant advantage in that it can be stored as a liquid at ambient temperature vs. liquid air, which must be stored at cryogenic temperatures between –190°C and –200°C.18
CO2 Battery is a technology concept being developed by Energy Dome, which uses a motor generator unit with a gas compressor and turbine to move CO2 between a low-pressure (ambient) storage dome and high-pressure storage cylinders to store and release energy.19 CO2 is compressed, condensed and stored as a liquid at ambient temperature and high pressure (but below its critical pressure) when excess power is available. The high-pressure liquid CO2 is vaporized, superheated and passed through a turbine to atmospheric pressure, generating power when energy is required to be released. Note: given that an approach is required between the ambient air and the CO2 in the condenser, condensation by air cooling will not be possible in most climates. Instead, refrigeration will be required, or, as will be shown later in this article, cold liquid can be generated by flashing supercritical CO2.
FIG. 11 shows a simplified sketch for a concept of a CO2 export compressor with an integrated liquid CO2 energy storage system. When excess power is available, CO2 from low-pressure storage can be combined with CO2 from the carbon capture plant and compressed up to an intermediate pressure for export to a CO2 gathering system. A portion of the CO2 from the discharge of the export compressor can be further compressed to 70 barg, and then condensed and stored as a liquid. When power is required, the liquid CO2 can be vaporized, superheated and passed through the turbine generator. The CO2 is let down in two stages, with superheating between the first and second turbines.
The export pressure considered is based on a typical CO2 gathering and offshore injection pipeline network in the UK (as depicted in FIG. 12), in which the gathering system runs at an intermediate pressure in the region of 40 barg, and the CO2 is compressed into the supercritical region at a booster station before being sent offshore for injection. For scenarios in which CO2 is transmitted as a dense-phase or supercritical fluid, the export pressure will likely be in the region of 100 barg, thereby eliminating the requirement for additional compressor stages to generate liquid CO2.
This concept is most suited to operations where the charging time is long vs. the discharge time. For example, consider two storage cycles: one in which the system charges for 20 hr and then discharges for 4 hr, and the other where the system charges for 4 hr and discharges for 20 hr. This example will consider a compressor that takes 250 MMsft3d of CO2 from a carbon capture plant at 0 barg and compresses it to 40 barg for export to a CO2 gathering system. The energy storage compressor will increase the pressure of a slipstream of the export CO2 to 70 barg, after which it will be condensed and stored as a liquid at its saturation temperature at 69 barg. The system will be sized to store and release 20 MWh of energy.
TABLE 7 shows that the overall energy storage efficiency is the same for both cycles. However, for the “quick discharge” case, the compressor size is only increased by 4% vs. 19% for the “slow discharge” case. This is because the extra energy that must be used to charge the CO2 storage system—which is the same for both cases—is spread over a longer charging period. Therefore, the capital investment required for the “quick discharge” option will be significantly lower than for the “slow discharge” option.
The low-pressure storage challenge. The requirement for low-pressure storage is the key disadvantage of this concept vs. LAES, which can utilize an effectively infinite low-pressure reservoir: the atmosphere. The 20-MWh example above required 383 t of CO2 to be stored at atmospheric pressure, with a storage volume of ~210,000 m3. The low-pressure storage volume could be reduced by storing CO2 at a slightly positive pressure (increasing low-pressure CO2 density); however, this may significantly increase equipment costs and reduce cycle efficiency. The mass of CO2 required for a given energy storage capacity can also be reduced by improving the efficiency of the process, such as by heating to a higher temperature at the inlet to each turbine stage.
Energy Dome proposed the use of flexible storage domes for low-pressure CO2 storage. The 20-MWh example above would require 30 hemispherical domes, each of them 30 m in diameter, to store the required volume of CO2. This would require a total plot area of 27,000 m2.. Other suppliers of double-membrane gas holders20 advertise storage capacities of up to 5,500 m3, which would provide a similar equipment count and required footprint.21,22
One potential alternative low-pressure storage option is a gas holder, also known as a gasometer (FIG. 13). These are large storage cylinders with movable caps that allow the storage volume to vary as mass is added or removed. Typical volumes for large gasometers are around 50,000 m3, with the largest ever built having a capacity of around 600,000 m3.23 Low-pressure storage for the 20-MWh case study could be provided by four gasometers measuring 50 m in diameter x 25 m in height, with a total plot area of 10,000 m2.
Both technologies use variable volume storage to maintain constant pressure regardless of the mass stored. If a fixed storage volume was used, the pressure would increase as mass was added to the system from the turbines. In such a configuration, the system would reach a settle-out condition where the inlet and outlet pressures of the turbine were balanced at some elevated pressure, limiting the amount of energy that could be extracted from the CO2. It would be challenging to achieve near-atmospheric storage unless extremely large storage volumes were used.
Any low-pressure storage option will require very large equipment; therefore, this presents a major challenge with this technology. By comparison, high-pressure storage is relatively compact and easy to achieve. For the 20-MWh case study, the 688 m3 of high-pressure storage can be provided by a single bullet of approximately 6.5 m diameter x 19 m in length.
Supercritical CO2 for energy storage. The following section investigates whether a similar concept is viable: using storage of CO2 as a supercritical fluid rather than a subcritical liquid. When excess power is available, or when energy prices are relatively low, additional power can be used to compress supercritical CO2 from the injection pressure up to an even higher storage pressure. When power demands or energy prices are high, CO2 from storage can be let down through a turbine to generate power. Assuming the same pressure profile for a CO2 storage and injection network like the one shown in FIG. 12, such a storage system would be best suited at a booster station where the CO2 is compressed into the supercritical region.
The key difference between this concept and the LAES or Energy Dome technology is the availability of a low-pressure reservoir. In this system (FIG. 14), the low-pressure reservoir is the injection pathway that will itself be at very high pressure. Because the high-pressure storage contains CO2 as a supercritical fluid, the storage pressure will decrease as mass is let down through the turbine. The pressure will decrease until it is balanced with the injection pressure. This is another key difference to the liquid CO2 energy storage system in which the inlet pressure to the turbine can remain constant, even as mass is released from the high-pressure storage.
With the liquid CO2 option, the CO2 is stored at its vapor pressure in high-pressure storage vessels. Therefore, the high-pressure storage pressure will remain relatively constant as mass is released from the system, with the only changes in pressure due to Joule-Thomson (J-T) cooling. In this design, the high-pressure storage is at supercritical conditions; therefore, the pressure will reduce as mass is released from the system and the density reduces.
The differential pressure (ΔP) across the turbine will be a function of time (Eq. 1). At a constant mass rate, and assuming constant turbine efficiency, the power generated (Q) will be a function of differential pressure (Eq. 2); therefore, the power generated will also be a function of time (t) (Eq. 3):
∆P = f(t) (1)
Q = f(∆P) (2)
Q = f(t) (3)
The relationship between Q and t can be determined numerically by splitting the process into small time steps; evaluating the density, differential pressure and power generated in each step; and then plotting a regression through the results. The relationship can be regressed by a quadratic equation in t.
FIG. 15 shows generic curves for power generation and high-pressure storage pressure against time. The total energy released is equivalent to the area under the power (red) curve.
The total time taken (T) for the high-pressure storage pressure to drop from its maximum value to the injection pressure can be expressed as shown in Eq. 4, where V is the storage volume (m3), m is the mass flowrate released from high-pressure storage through the turbine (kg hr-1), and Δρ is the difference in density of the CO2 at its high-pressure storage and injection pressures (kg m-3):
T = VΔρ / m (4)
The total amount of energy (E) released during the time taken for high-pressure storage pressure to fall from its initial value until it balances with the injection pressure can be calculated, as shown in Eq. 5:
E = (m / V) ∫ 0T Q dt (5)
FIG. 16 shows the results of a case study where the energy released per m3 of high-pressure storage volume was calculated for a range of high-pressure storage and injection pressures, with the fluid heated to 150°C at the turbine inlet.
This demonstrates the key drawback of this option—the energy storage density is an order of magnitude lower than the liquid CO2 storage concept. This is a result of the relative lack of expansion across the turbine compared to that which results from passing gas-phase CO2 through the turbine from high pressure down to low (near-atmospheric) pressure, as is the case with the liquid CO2 storage option. This results in significantly less differential enthalpy across the turbine, and, thus, decreased power output.
This can be seen graphically by plotting the turbine paths for each option on a pressure-enthalpy diagram (FIG. 17). For the supercritical storage option, the enthalpy change is 1,491 kJ mol-1, and the wide spacing of the isotherms in this region means there is a small change in temperature across the turbine. For the vaporized liquid storage option, the enthalpy changes across the first and second stages of the turbine are 3,528 kJ mol-1 and 4,717 kJ mol-1, respectively, resulting in a total enthalpy change of 8,246 kJ mol-1. The isotherms in this region are very compact, meaning that there is a significant drop in temperature across each stage, and that significant enthalpy can be put into the system in the interstage heater using a relatively low-grade heating medium—in this case, at 150°C.
This difference—combined with the fact that, for the supercritical option, the power generation continuously drops as mass is released from high-pressure storage—means that the energy storage density of the supercritical option over an entire discharge cycle is significantly less than the liquid storage option.
Concept for a propylene carbonate carbon capture plant with integrated refrigeration and energy storage systems. This section outlines a conceptual design for an autothermal reformer (ATR), pre-combustion carbon capture plant using propylene carbonate, injection compression with an integrated CO2 refrigeration loop and energy storage. The plant will produce 150,000 tpy of 99.9%-purity H2. The CO2 compressor increases the pressure to 180 barg for injection. FIG. 18 shows a simple process flow diagram of the process. TABLE 8 summarizes key process data for the plant.
Refrigeration. For further details, refer to Part 1 of this article. The chiller duty is provided using recycle around the CO2 injection compressor, from the discharge of the sixth stage at 120 barg back to the suction of the third stage at 9.7 barg. The recycle rate required is 123 tph, providing an additional compressor power of 5.8 MW. For comparison, a closed-loop CO2 refrigeration system to provide this duty would have a compressor power of 4.9 MW.
Energy storage. The energy storage system uses CO2 from the discharge of the sixth stage at 120 barg. The CO2 at this stage is supercritical, and liquid CO2 is produced by flashing this stream down to 70 barg and storing the liquid at its vapor pressure in the high-pressure storage vessel(s). The flashed stream has a vapor fraction of 0.09; therefore, this vapor is returned to the suction of the sixth stage at 70 barg. This configuration is an improvement over condensing CO2 at 70 barg, as it avoids the requirement for further refrigeration duty to condense the gas phase and produce the liquid CO2 for storage. The design of the sixth-stage aftercooler is very important, as the vapor fraction of the flashed stream is very sensitive to the temperature upstream of the J-T valve.
A large portion of energy generated by the turbines is sourced from the heat input to the heaters at the inlet of each turbine stage. Therefore, it is highly desirable if this energy can be sourced from otherwise unused waste heat. One potential source is the outlet of the ATR, which is at very high temperature (around 1,000°C). Much of this heat is recovered by raising steam in waste heat boilers, and then various other exchangers for heat integration. The final stream that leaves these exchangers is at around 160°C; therefore, it retains significant low-grade heat, which is typically rejected into the atmosphere through air cooling. In this case study, the air cooler at the ATR plant outlet has a duty of 66 MW. While a lot of heat duty is available, the limiting factor is the approach temperature between the process side and the energy storage side. A negative hot-out/cold-out (HOCO) approach is possible, but usually requires multiple exchanger shells arranged in series, which can have a significant impact on capital cost and plot space.
FIG. 19 shows the results of a case study considering different amounts of energy storage, generalized per MW of Base Case (no energy storage capability) compressor power (in this case, 20.9 MW). All points have the same energy storage efficiency of 45%. This is lower than the earlier example because this configuration accounts for additional compressor recycle as a result of vapor generation through the J-T valve, and thus additional energy input to generate the liquid CO2.
A storage capacity of 1 MWh–2 MWh per MW of compressor duty offers a reasonable balance between compressor duty and storage capacity. In this example, a storage capacity of 20 MWh was selected with a 20-hr/4-hr charge/discharge cycle. This requires 2.24 MW (equivalent to 11%) additional compressor power and gives an HOCO approach of 4°C in the heaters, which can be accommodated in a single heat exchanger shell. Results for this case study are shown in TABLE 9.
For this scheme to be economical, the difference in energy price between peak and non-peak periods must be such that the cost of the additional energy used during the charge sequence is offset by the power generated during the discharge sequence. This will be the case when the following expression is true (Eq. 6):
(Pd / Pc ) ≥ (Tc / Td ) [(Qc – Qd ) / Gd ] (6)
The right-hand side of Eq. 7 is the inverse of the energy storage efficiency:
ηE = (Td / Tc ) [Gd / (Qc – Qd )] (7)
Therefore, an energy storage system can only be viable if Eq. 8 is true:
ηE ≥ Pc / Pd (8)
where:
ηE = Energy storage efficiency
Pd = Energy price during discharge (£ MWh-1)
Pc = Energy price during charge (£ MWh-1)
Tc = Charge time, hr
Td = Discharge time, hr
Qc = Compressor power during charge, MW
Qd = Compressor power during discharge, MW
Gd = Power generated during discharge, MW
For this example, the energy efficiency is 45%; therefore, the price ratio must be less than 0.45. FIG. 20 shows the off-peak-to-peak price ratio in the UK for the period of May 2019–December 2022. The average for this period is 0.49, with values ranging from 0.35–0.7.14
The peak period is defined here as the amount of time that the energy price is more than 20% over the daily average. This corresponds to an average daily peak length of 3.6 hr. The viability of this concept is highly dependent on the achievable energy storage efficiency. This can be greatly improved if higher-grade waste heat sources are available, as this would enable higher turbine preheat temperatures and an associated increase in energy generation with no increase in compressor power. Energy Dome claims a storage density of 66.7 kWh m-3 by capturing and storing the heat of compression and then reusing this energy for heat input upstream of the turbines.19 One potential avenue for further study is to optimize the heat integration between the ATR and CO2 energy storage units.
Takeaways. Some typical parameters of energy storage systems discussed in this article are detailed in TABLE 10. Liquid CO2 has potential as an alternative energy storage medium over similar technologies such as compressed air energy storage (CAES) and LAES. The key downside is the requirement for low-pressure storage; however, there is the benefit of having no requirement for low-temperature storage.
The viability of a liquid-CO2 energy storage system is strongly dependent on the availability of waste heat streams and the energy storage efficiency. The case study presented in this article, which considers recovering waste heat from the outlet of an autothermal reformer plant, yields an energy storage density of 29 kWhm-3. Energy Dome claims an energy storage density of 66.7 kWhm-3 by storing and reusing waste heat of compression.19 If this claimed energy storage density could be realized on a full-scale plant, it would be an order of magnitude higher than CAES but would still fall short of LAES or a lithium-ion battery.
Supercritical CO2 energy storage is a less-attractive solution for two main reasons. First, storage as a supercritical fluid results in a reduction in high-pressure storage pressure when material is released through the turbine, significantly limiting the energy storage capacity. Second, the relative lack of expansion and differential enthalpy across the turbine results in poor energy efficiency. However, flashing supercritical CO2 to generate the liquid for storage is a significant improvement over condensing gas-phase CO2 at 70 barg, as it avoids the requirement for additional refrigeration duty. Therefore, the integration of a liquid-CO2 energy storage system with a supercritical CO2 injection system presents opportunities for improved efficiency.
A standalone liquid-CO2 energy storage system is viable if gas phase is being exported into a network at some intermediate pressure that is below the critical pressure (e.g., 40 barg); however, this will require some form of external refrigeration to provide condensing duty. If the captured CO2 is being compressed up to a high injection pressure, there is the possibility to integrate the refrigeration system with the export compressor and generate liquid CO2 by flashing the supercritical fluid. Therefore, an ideal application is a propylene carbonate carbon capture system where the captured CO2 is being compressed for injection. This enables the solvent chilling loop and liquid energy storage systems to be integrated with the export compressor, eliminating the requirement for any external refrigeration systems.
The economic viability is questionable unless favorable electricity pricing incentives are implemented. The energy price ratio must be such that the cost of the additional energy used during the charge sequence is offset by the power generated during the discharge sequence. This can be improved significantly if greater energy storage efficiency can be achieved by utilizing sources of waste heat. HP
ACKNOWLEDGMENTS
The authors would like to thank Adish Jain, Process Technology Director, Fluor, and Samantha Nicholson, Process Fellow, Fluor, for their valuable input to this article.
LITERATURE CITED
CHARLIE GOULD is a Process Engineer at Fluor. He has 8 yr of experience on a variety of upstream oil and gas, refining, chemicals, heat and power, carbon capture and hydrogen projects. He specializes in steady-state and dynamic process simulation, overpressure protection and process safety time analysis. Gould earned an MEng degree in chemical engineering from the University of Surrey in the UK and is a chartered engineer with IChemE.
ROBERT JOYNER is a Principal Process Engineer at Fluor. He has more than 20 yr of experience in onshore and offshore oil and gas, refining and chemicals projects, specializing in overpressure protection and flare system design. He earned an MEng degree in chemical engineering from Imperial College London and is a chartered engineer with IChemE.
ADYA DESHMUKH is a Senior Process Engineer at Fluor. She has more than 20 yr of experience, specializing in CO2 processing, including Econamine FG PlusSM, direct air capture and CO2 compression/conditioning process design. She graduated from the University of Surrey in chemical engineering and is a registered chartered engineer.