3 Assessment of operating flexibility of power plants with CCS

The reference plants selected for the assessments of this study are the NGCC, IGCC, USC PC and Oxy-combustion plant. For the combined cycle-based alternatives (NGCC and IGCC), the design capacity of the plant is fixed to match the appetite (thermal requirement) of two F-class gas turbines at the reference ambient temperature of the study (9°C). For the boiler-based alternatives (USC PC and Oxy-combustion plant), the design capacity is selected by referring to a boiler size that could be currently engineered and built, corresponding to approximately 750-1000 MWe gross power production.

The economic data of each case have been derived from the data contained in the reference studies, after currency adjustment and cost level escalation.

For the reference plants with leading CCS technologies, the following sections identify the elements that may constrain the operating flexibility of the plant, discuss possible ways of overcoming them and assess performance and cost implications of flexible operation. Some elements are common to the different power plant types, while others are related to a specific technology only.

Figure 3-1: Load operation of power plants with CCS

Image

Depending on the power plant type, these considerations are based on the assumption that plants will be requested to operate in the mid and peak merit market, in order to meet actual power market requirements. The trends assumed for the different power plants follow a weekly demand curve characterised by two operating regimes, as shown in Figure 3-1. Additional considerations have been made by considering alternative scenarios, as explained in the following:

  • A weekly demand curve characterised by three operating regimes, with two hours per working day of peak electricity demand, as shown in Figure 3-2.
  • An electricity market where the USCPC plant and the power train of the IGCC are shutdown analogously to the demand curve of the combined cycles.

Figure 3-2: Three regimes load operation of power plants with CCS

Image

3.1 Thermal cycling of power plants with CCS

In general, the introduction of the CO2 capture and compression facilities in power plants may impose additional constraints to a flexible operation, predominantly for the combined cycles and the USCPC plants with post-combustion capture, where certain equipment, like stripper and reboiler, may limit the capacity to make frequent start-ups/shut-downs, due to the time required to pre-heat the regeneration column and the related reboilers. For plants with other capture technologies, i.e. pre-combustion capture and cryogenic purification of oxy-combusted flue gases, this constraint is not present as the capture unit is generally capable to follow the transient operation of the other units.

For the NGCC and USCPC plants, to overcome this constraint it is possible to consider the storage of CO2-laden solvent (Case 1a and 3a), which allows to decouple the Gas Turbine or the boiler island from the CO2 capture unit during startup. As an alternative, a small fired heater providing the heat required for preheating the regenerator column before the plant start-up could be installed, avoiding the need for solvent storage during this phase. However, with this solution a certain amount of CO2 in the flue gas from the fired heater is released to the atmosphere.

Recently designed combined cycle plants can be started-up in 45-55 minutes, after night shutdown (hot start-up), or 2 hours after weekend shutdown (warm start-up), while recently designed USC PC plants can be started-up respectively in 120 minutes and less than 4 hours. On the other hand, the heating up of a regenerator column could require a few hours, once the steam is available from the steam cycle. In this case, solvent circulation in the CO2 absorber can be started before gas turbine/boiler ignition so that, when gas turbine/boiler is started-up with its own ramp-up rate, the exhaust gases are fed to the absorption column and CO2 is captured by lean solvent. As soon as steam from the HRSG/boiler is available at required pressure, the regeneration section can be heated up. It has been estimated that the regeneration section can be ready for operation at full load in 120 minutes, after gas turbine/boiler ignition during hot start-up, while 240 minutes are required in case of warm start-up. In order not to limit the operating flexibility of the combined cycle with CCS, the strategy considered in Case 1a and 3a is that until the regenerator is not able to purify the CO2-rich amine from the bottom of the absorber, rich solvent is sent to a storage tank, while lean amine and semi-lean amine are taken from other dedicated tanks.

The solid lines in Figure 3.1-1 show for Case 1a the solvent flowrate from/to the storage tanks during hot start-up, while the dashed lines represent the resulting required storage volume (similar trend is during warm start-up).

Figure 3.1-1: Case 1a (NGCC) – Stored solvent volume during hot start-up

Image

For the NGCC case, two alternatives have been assessed for the regeneration of stored rich solvent and refilling of lean and semi-lean amine storage tanks:

  1. Regeneration during off-peak hours, maintaining the plant in operation at minimum environmental load, i.e. one gas turbine operated at about 40%, for approximately 3-4 hours per night in order to provide steam for the reboiler.
  2. Regeneration during peak hours, when the plant is operated at full load, thus requiring an oversize of about 15% for the regeneration and compression units.

The first alternative is considered the most reasonable choice, because it has the lowest investment cost and the highest power production during peak demand period. However, higher variable and fixed operating costs will need to be considered during off-peak demand period, because the power plant is operated at minimum environmental load for the time required to regenerate rich solvent and refill lean solvent tanks. Figure 3.1-2 shows the dynamic trend of the stored solvent volume during the week. The design of the storage tanks is fixed by the amount of stored solvent required during warm start-up.

Figure 3.1-2: Case 1a (NGCC) – Stored solvent volume during the week

Image

For the USCPC plant following a two regimes demand curve where the plant is required to be shutdown during low electricity demand period (Case 3a – Scenario 2), the regeneration of stored rich solvent and refilling of lean and semi-lean amine storage tanks is carried out when the plant is operated at full load, thus requiring an oversize of about 8.5% for the regeneration and compression units.

Figure 3.1-3 shows the dynamic trend of the stored solvent volume during the week. The design of the storage tanks is fixed by the amount of stored solvent required during warm start-up.

For the USCPC plant following a three regimes demand curve (Case 3a – Scenario 3), where the plant is shutdown during low electricity demand period and to cover two hours per working day of peak electricity demand, the regeneration of stored rich solvent and refilling of lean and semi-lean amine storage tanks is carried out during normal electricity demand, thus requiring an oversize of about 24% for the regeneration and compression units. Figure 3.1-4 shows the dynamic trend of the stored solvent volume during the week. The design of the storage tanks is fixed by the amount of solvent stored after peak demand period on Monday.

Figure 3.1-3: Case 3a (USC PC plant) - Scenario 2 - Stored solvent volume during the week

Image

Figure 3.1-4: Case 3a (USC PC plant) - Scenario 3 - Stored solvent volume during the week

Image

The following table summarizes the main performance and cost data of Case 1a and 3a (Scenario 2 and 3).

Table 3.1-1: Thermal cycling in NGCC – Performance and cost data summary (Est. accuracy: ±35%)

Tag Plant type Reference plant Regeneration during off-peak
Performance TIC, M€ Performance Size (% of ref. plant) / Plant changes TIC, M€
Case 1a NGCC w post-comb NPO=742MWeNEE=50.6% 726 PeakNPO=742MWeNEE=50.6% Start-up ST65MWeCondensing section 190% 783
Off-peak (during regeneration)NPO=77MWeNEE=18.4% Rich solvent 2 × 12,500 m3 (D: 31.1 m × H: 16.5 m)
Lean solvent 1 × 13,000 m3 (D: 31.1 m × H: 17.1 m)
Semi Lean solv: 1 × 12,500 m3 (D: 31.1 m × H: 16.5 m)
Case 3a (Scenario 2) USC PC w post-comb NPO=666 MWeNEE=34.8% 1,513 PeakNPO=655MWeNEE=34.2% Regeneration / compression section 108.5% 1,545
(Plant shutdown during off-peak) Rich solvent 2 × 12,000 m3 (D: 30.5 m × H: 16.5 m)
Lean solvent 1 × 13,000 m3 (D: 31.1 m × H: 17.1 m)
Semi Lean solv: 1 × 12,000 m3 (D: 30.5 m × H: 16.5 m)
Case 3a (Scenario 3) USC PC w post-comb NPO=666MWe NEE=34.8% 1,513 PeakNPO=808MWeNEE=42.2% Regeneration / compression section 124% 1,627
New ST: 113 MWe Condensing section 145%
Normal operationNPO=655MWeNEE=34.2% Rich solvent 2 × 17,300 m3 (D: 36.6 m × H: 16.5 m)
(Plant shutdown during off-peak) Lean solvent 1 × 17,300 m3 (D: 36.6 m × H: 16.5 m)
Semi Lean solv: 1 × 17,300 m3 (D: 36.6 m × H: 16.5 m)

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost Estimate accuracy: ±35%

It can be drawn that for power plants with CO2 post-combustion capture, to maintain same thermal cycling capability as the conventional plants without capture, solvent storage is required, leading to an investment cost increase of about 8% and 2% with respect to the reference case, respectively for NGCC and USCPC boiler cases, considering a weekly demand curve with two operating regimes.

A higher investment cost, around 7.5% of the reference case, is required for the USCPC boiler case, when considering a weekly demand curve with three operating regimes.

3.2 CO2 capture solvent storage

For NGCC and USC-PC power plants, the introduction of the post-combustion solvent washing process and the CO2 compression unit may potentially limit their intrinsic capacity to operate flexibly. However, solvent storage can allow to decouple the operation of the absorption section from the regeneration and compression units, while continuously capturing the CO2 from the flue gases. Solvent regeneration and compression, with their associated energy penalties, can then be made during low electricity demand periods. This feature has the potential for improving load following capabilities and overall economics of capture plants, because the electricity production can be maximized when the market requires a higher electricity generation.

Licensors of the most referenced solvent washing technologies, like Aker Clean Carbon, Alstom and Mitsubishi Heavy Industries have all confirmed the technical feasibility of solvent storage, either lean or laden, provided that the temperature of the rich solvent is maintained at or slightly below absorber bottom outlet temperature condition, to avoid degassing or venting of carbon dioxide and potential over pressure of the tank. Furthermore, high rates of solvent degradation in the rich storage tank are not expected; degradation would be mainly due to the reaction with oxygen, therefore nitrogen or CO2 blanketing shall always be considered. In addition, solvent solution is not flammable at the concentration used in the capture plant and cannot be auto-ignited during different operating modes.

Furthermore, MHI owns a patent in the European Union, USA and Japan (EP 0537593B1), which is dedicated to the storing of solvent and regeneration during high power demand.

3.2.1 Solvent storage for plants with two operating regimes

Cases 1b (NGCC plant) and 3b (USCPC plant) are based on a weekly demand curve characterized by two operating regimes, as shown in Figure 3-1. For these plants, to maximize energy production, rich solvent can be partially or even totally stored during the 80 hours per week of peak load operation, when the plant is at base-load, while the regeneration of stored solvent can be made during the remaining 88 hours per week of off-peak load operation, when the plant is required to operate at a partial load (50% NPO for USC-PC) or is shutdown (NGCC). With this strategy, the solvent flowrates from/to the storage have to be balanced in one week of operation.

During peak electricity demand, when the market requires maximum amount of electricity, the power plant is operated at base load by making full capture of the CO2 from the flue gases in the absorber column, while only a certain amount of the CO2-rich solvent from the absorber bottom is fed to the regenerator, the remainder being stored in dedicated storage tanks. As a consequence, part of the lean and semi-lean solvent required for the CO2 capture in the absorber is not available from the regenerator, so it is has to be taken from dedicated storage tanks.

During off-peak electricity demand, i.e. when lower electricity selling prices reduce the revenues of the plant, the stripper can be operated in order to regenerate the rich solvent stored in the tanks, while refilling the lean amine storage tanks. The steam required for the regeneration is taken from the power island, thus implying that the combined cycle has to be operated at minimum environmental load, i.e. the shutdown required by the electricity demand curve is not possible for this plant type.

Different regeneration loads during high electricity demand period have been investigated in order to evaluate the most convenient operating condition. The resulting optimum regeneration loads are 50% and 25%, respectively for NGCC and USC-PC power plants, thus resulting in a significant increase of the net power output during peak hours, while avoiding the need for excessive storage volumes. For each plant, two possible scenarios have been considered:1) Reduced (i.e. lower than reference plant) size of the regeneration and compression section, resulting in 74% and 85% of the reference case, respectively for the NGCC and the USC-PC; 2) Same size as the reference plant, i.e. unchanged design.

Figure 3.2-1 shows the stored volumes of solvents during the week, for the scenarios considered in the NGCC plant (same trend is for the USC-PC case). The net volume of the storage tank is the difference between the maximum and the minimum volume of solvent stored during the week. It corresponds to the solvent stored during the weekend, from turndown of Friday night to ramp up of Monday morning. The solid line corresponds to the stored volume for scenario 1, while the dashed line corresponds to the stored volume for scenario 2. Although both scenarios are designed for the same regeneration load during peak time, storage tanks required for the second alternative are smaller because it is possible to maintain this section at base load during off-peak hours of the working days, while maintaining a lower load during the week-end, enough to avoid accumulation in the storage tanks.

Figure 3.2-1: NGCC –Stored solvent volume during the week

Image

The following tables summarize the main performance and cost data of the two power plants. From the figures in the tables the following conclusions can be drawn:

  • By introducing adequate solvent storage in the plant, the electricity production and the net electrical efficiency during peak demand period increase by about 5% to 6% with respect to the reference case.
  • For the NGCC plant, the investment cost delta is about 20% higher than the reference case, both for the alternative with reduced regeneration and compression units design and the case with unchanged design. Cost delta variation for the USC-PC plant with respect to the reference plant is respectively 7%.
  • When comparing the two alternatives, it follows that an unchanged design (scenario 2) is the most attractive choice. In fact, this alternative has both a wider operating flexibility and a slightly lower investment cost.

Table 3.2-1: Scenario 1 (lower size) – Performance and cost data summary (Estimate accuracy: ±35%)

Tag Plant type Reference plant Scenario 1 (lower size)
Performance TIC, M€ Performance Storage tanks TIC, M€
Case 1b NGCC w post-comb NPO=742 MWeNEE=50.6% 726 NPO=788 MWeNEE=53.7% Rich solvent 2 × 87,500 m3 (D: 81 m × H: 17 m) 885
Lean solvent: 1 × 87,500 m3 (D: 81 m × H: 17 m)
Semi Lean solvent: 1 × 87,500 m3 (D: 81 m × H: 17 m)
Case 3b USC PC w post-comb NPO=666 MWeNEE=34.8% 1,513 NPO=697 MWeNEE=36.4% Rich solvent 2 × 71,600 m3 (D: 73 m × H: 17 m) 1,627
Lean solvent: 1 × 71,600 m3 (D: 73 m × H: 17 m)
Semi Lean solvent: 1 × 63,600 m3 (D: 69 m × H: 17 m)

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost

Table 3.2-2: Scenario 2 (full size) - Performance and cost data summary (Estimate accuracy: ±35%)

Tag Plant type Reference plant Scenario 2 (full size)
Performance TIC, M€ Performance Storage tanks TIC, M€
Case 1b NGCC w post-comb NPO=742 MWeNEE=50.6% 726 NPO=788 MWeNEE=53.7% Rich solvent 2 × 71,600 m3 (D: 73 m × H: 17 m) 868
Lean solvent: 1 × 71,600 m3 (D: 73 m × H: 17 m)
Semi Lean solvent: 1 × 71,600 m3 (D: 73 m × H: 17 m)
Case 3b USC PC w post-comb NPO=666 MWeNEE=34.8% 1,513 NPO=697 MWeNEE=36.4% Rich solvent 2 × 47,700 m3 (D: 60 m × H: 17 m) 1,605
Lean solvent: 1 × 55,700 m3 (D: 65 m × H: 17 m)
Semi Lean solvent: 1 × 47,700 m3 (D: 60 m × H: 17 m)

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost.

3.2.2 Solvent storage for plants with three operating regimes

Cases 1f (NGCC plant) and 3e (USCPC plant) are based on a weekly demand curve characterized by the following three operating regimes:

  • Peak electricity demand period: 2 hours per working day.
  • Normal operation: 14 hours per working day.
  • Off-peak electricity demand period (NGCC plant shutdown or USC PC generating 50% of net power output): night and weekend.

To maximize the energy production, the rich solvent is totally stored during the 2 hours per day of peak load operation, when either the gas turbines or the boiler are at 100%load. The power plant is operated at base load by making the full capture of the CO2 from the flue gas in the absorber column, while the solvent regeneration and CO2 compression sections are halted. A supplementary LP steam turbine has been considered to expand the additional steam available when the regeneration is halted; this avoided to over sizing the steam turbine for the total amount of steam, as well as the inefficient operation of the machine during normal operation.

For the NGCC case, as per the assumed electricity demand curve, the plant is fully shut down overnight and at the weekend, while the regeneration of stored solvent is made during the 14 hours per day of normal operation, thus requiring an oversize of the regeneration and compression section of approximately 14% to avoid any accumulation of the stored solvent.

For the USCPC case, the regeneration of stored solvent can be made during the 8 night hours per day of off-peak load operation, when the plant is required to operate at a partial load in order to produce 50% of the normal operation net production. This leads to a boiler load around 55% during the weekend and 61% during weekday night time, when the solvent stored during peak load operation has to be regenerated, while the regenerator and compression section operate at around 86%.

With this strategy, the solvent flowrates from and to the storage are balanced within each day of plant operation, leading to a size of the storage tanks that is smaller than the demand curve based on two operating regimes, as shown in the previous section.

The following tables summarize the main performance and cost data of the two power plants. From the figures in the tables the following conclusions can be drawn:

  • By introducing adequate solvent storage in the plant, the electricity production and the net electrical efficiency during peak demand period increase from about 12% to 22% with respect to the reference case. For the NGCC plant, during normal operation the net power output is around 2% lower than the reference case, due to the oversize of the regenerator, which also corresponds to an increased pipeline diameter (400 mm vs. 350 mm)
  • For the NGCC plant, the investment cost delta is about 9% higher than the reference case. Cost delta variation for the USC-PC plant is 6%.

Table 3.2-3: Daily cycle solvent storage – Performance and cost data summary

Tag Plant type Reference plant Daily cycle solvent storage with an alternate demand curve
Performance/ pipe diam. (mm) TIC, M€ Performance pipe diam. (mm) Size (% of ref. plant) / Plant changes TIC, M€
Case 1f NGCC w post-comb capture NPO=742MWeNEE=50.6%\ 726 PeakNPO=832MWeNEE=56.7% Regeneration / compression section 114% 793
Pipeline D: 350 100km pipe: 167 Normal operationNPO=729MWeNEE=49.6% New ST: 77MWe Condensing section 195% 100km pipe: 185
Pipeline D: 400 Rich solvent 2 × 7,600 m3 (D: 27.4 m × H: 12.8 m)
Lean solvent: 1 × 7,600 m3 (D: 27.4 m × H: 12.8 m)
Semi Lean solvent: 1 × 7,600 m3 (D: 27.4 m × H: 12.8 m)
Case 3e USCPC w post-comb capture NPO=666 MWeNEE=34.8% 1,513 PeakNPO=813 MWeNEE=42.5% New ST: 91MWe New condenser 295 MWth 1,600
Normal operationNPO=666 MWeNEE=34.8% Rich solvent 2 × 12,000 m3 (D: 30.5 m × H: 16.5 m)
Lean solvent: 1 × 12,000 m3 (D: 30.5 m × H: 16.5 m)
Semi Lean solvent: 1 × 10,100 m3 (D: 27.4 m × H: 17 m)

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost Estimate accuracy: ±35%

3.3 Constant CO2 flowrate in transport pipeline

For each power plant assessed in the study, the cycling operation required to meet the variable grid demand leads to an uneven captured CO2 flowrate and a consequent fluctuation of the operating conditions in the pipeline. As a consequence, a two-phase flow or a significant change of the physical properties could occur in the pipeline, if pressure and temperature were not maintained within a limited range of variation with respect to the normal operation of the capture plant. Furthermore, for some applications like the Enhanced Oil Recovery (EOR) it would be preferred to have a pre-determined flow rate of CO2, even if variable, rather than an unpredictable fluctuating stream. Two different options have been considered to avoid these issues:

  • Scenario 1 (CO2 buffer storage): introduction of a CO2 storage system, to maintain a constant CO2 flowrate in the pipeline.
  • Scenario 2 (Reduced regenerator capacity, valid for post-combustion technologies): operation of the regeneration and compression sections at constant and reduced load. These sections are designed for a lower capacity, while solvent storage tanks compensate the difference between the absorber and the regenerator load.

Using above strategies, a constant CO2 flowrate lower than peak production when the plant is operated at base load is sent to the external pipeline; then, it is possible to select a lower pipeline diameter, leading to a potential cost saving, depending on the overall length of the pipeline, though some costs associated with laying a pipe (e.g. access, earthmoving) are generally more dependent on length, rather than diameter.

For Scenario 1, Figure 3.3-1 shows a trend, typical for all plant types, of the whole volume of stored CO2 during the week and the single vessel volume trend. The required net volume of the storage vessels is the difference between the maximum and the minimum volume of stored CO2 during the week. From the graph, it can be drawn that it corresponds to the CO2 accumulated during the weekdays and mainly discharged during the partial load operation from Friday night to Monday morning.

With reference to Scenario 2, Figure 3.3-2 shows a trend, typical for all plant types, of the stored volumes of rich, lean and semi-lean solvents during the week. The net volume of the storage tank corresponds to the difference between the maximum and the minimum volume of solvent stored during the week. It corresponds to the solvent stored during the weekend, from turndown of Friday night to ramp-up of Monday morning.

Table 3.3-1 and Table 3.3-2 summarize main performance and cost data of different plants. For each case, estimated cost of 100 km pipeline is also included in the figure.

Figure 3.3-1: Scenario 1 – Stored CO2 volume during the week

Image

Figure 3.3-2: Scenario 2 –Stored solvent volume during the week

Image

Table 3.3-1: Scenario 1 (CO2 buffer storage) – Performance and cost data summary

Tag Plant type Reference plant Scenario 1 (CO2 buffer storage)
Performance/ pipe diam. (mm) TIC, M€ Performance (peak hours)/ pipe diam. (mm) CO2 storage vessels TIC, M€
Case 1d NGCC w post-comb capture NPO=742 MWeNEE=50.6%Pipeline D: 350 726 100km pipe: 167 NPO=742MWeNEE=50.6%Pipeline D: 250 6×1,535 m3(D: 8.7m, H: 26.1m) 748 100km pipe: 135
Case 2e IGCC w pre-comb capture NPO=730 MWeNEE=31.4%Pipeline D: 500 1,885100kmpipe: 206 NPO=732 MWeNEE=31.4%Pipeline D: 450 8×1,600 m3(D: 8.8m, H: 26.4m) 1,915100kmpipe: 195
Case 3c USC PC w post-comb capture NPO=666 MWeNEE=34.8%Pipeline D: 500 1,513100kmpipe: 206 NPO=666 MWeNEE=34.8%Pipeline D: 450 6×1,450 m3(D: 8.5m, H: 25.5m) 1,541100kmpipe: 195
Case 4c Oxy-combustion USC PC NPO=533MWeNEE=35.5%Pipeline D: 500 1,387100kmpipe: 206 NPO=536MWeNEE=35.7%Pipeline D: 400 6×1,325 m3(D: 8.3m, H: 24.9m) 1,408100kmpipe: 184

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost Estimate accuracy: ±35%

Table 3.3-2: Scenario 2 (Lower regenerator/compressor size) – Performance and cost data summary

Tag Plant type Reference plant Scenario 2 (lower size)
Performance/ pipe diam. (mm) TIC, M€ Performance (peak hours)/ pipe diam. (mm) Size (% of ref. plant) / Storage tanks TIC, M€
Case 1d NGCC w post-comb capture NPO=742MWeNEE=50.6%Pipeline D: 350 726100kmpipe: 167 NPO=776MWeNEE=52.9%Pipeline D: 300 Regeneration section 62.5% 838 100km pipe: 150
Rich solvent 2 × 63,600 m3 (D: 69 m × H: 17 m)
Lean solvent: 1 × 63,600 m3 (D: 69 m × H: 17 m)
Semi Lean solvent: 1 × 63,600 m3 (D: 69 m × H: 17 m)
Case 3c USCPC w post-comb capture NPO=666 MWeNEE=34.8%Pipeline D: 500 1,513100kmpipe: 206 NPO=688 MWeNEE=36.0%Pipeline D: 450 Regeneration section 80% 1,601100kmpipe: 167
Rich solvent 2 × 55,700 m3 (D: 65 m × H: 17 m)
Lean solvent: 1 × 63,600 m3 (D: 69 m × H: 17 m)
Semi Lean solvent: 1 × 55,700 m3 (D: 65 m × H: 17 m)

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost Estimate accuracy: ±35%

From the figures in the tables the following conclusions can be drawn:

  • By introducing CO2 buffer storage in the plant, overall performances during peak time are basically not affected, while the total investment cost increase (not including the pipeline) is marginal, ranging from 2% to 3% of the reference case. However, depending on the overall length, this investment increase may be offset by the lower cost of the pipeline.
  • For the NGCC and the USC-PC alternatives, if solvent storage is introduced in the plant then the electricity production increases by about 3% to 5%, during peak hours, with respect to the reference case. On the other hand, the plant total investment cost is respectively 12% and 4% higher than the CO2 buffer storage option (overall % increase equal to 15.4 and 5.8 respectively).

3.4 Hydrogen storage in IGGC plants with CCS

The operating flexibility and economics of the IGCCs can be improved if the plant is designed for the co-production of electricity and hydrogen (Case 2b) or if a buffer storage of hydrogen rich gas (Case 2c and 2f) is introduced in the plant. In this case, the syngas (or hydrogen) production line and CCS plant can operate constantly at full load, while the hydrogen-fired power plant follows the requirements of the flexible market (i.e. demand curve with two operating regimes).

In all the alternatives assessed in the study, part of the hydrogen rich gas from the CO2 removal unit is fed to storage during low electricity demand periods, while it is used during electricity peak demand.

During low electricity demand period and for Cases 2b and 2c, the excess syngas production, obtained from the process units running at base load, is stored or used to produce hydrogen, while the power plant is operated with two gas turbines at their minimum environmental load, which is 60% of base production, corresponding to approximately 66% of fuel requirement. In Case 2f, as the plant is required to operate in island mode during off-peak demand period, only one gas turbine is in operation at its minimum environmental load.

For Case 2b, the amount of fuel required by the gas turbines is sent to the power island for electricity generation, while the remainder part from the AGR, corresponding to approximately 34% of the overall production, is split into two different streams: one is fed to a pressure swing adsorption (PSA) unit for high purity hydrogen production, while the other stream is sent to underground storage and used as feeding stream for the PSA during peak-hours operation, i.e. when all the syngas generated from the gasification island is dedicated to the power production. The PSA design capacity is selected to generate a constant hydrogen flowrate at plant battery limits, during the whole week of plant operation. It has been estimated that by storing approximately 48% of the de-carbonised fuel used for hydrogen production during off-peak demand period, then the PSA can be maintained at constant load, producing about 75,400 Nm3/h of high purity hydrogen.

For Case 2c and 2f, fuel gas from/to the storage system has to be balanced during the cyclic weekly operation, in order to avoid any accumulation of fuel. The need of balancing the fuel gas fixes the design capacity of the whole syngas generation line, which results in 82% and 65% of the reference case, respectively for Case 2c and 2f.

During high electricity demand period, the power island is operated with the two gas turbines at base load. For Case 2b, hydrogen rich gas from the storage is fed to the PSA to generate a constant hydrogen flow, while for Case 2c and 2f, where the process units are designed for a lower capacity, the hydrogen rich gas from the AGR unit is integrated with the stored gas, to meet the thermal requirement of the two machines.

It is noted that, as the ASU and the power trains are maintained at different loads during the cyclic operation, the air integration between the ASU and the gas turbines may potentially represent a constraint for the flexible operation of the IGCC. In this case, an additional main air compressor shall be considered for operation during off-peak hours, as the air extracted from the gas turbines, operated at part load, is significantly lower than the amount required by the ASU, operated at base load.

Figure 3.4-1 shows the main hydrogen rich fuel flowrate on the whole week of plant operation and the related volumes of stored gas for Case 2b. From the graph, it can be concluded that a storage volume of about 100,000 m3 is required for this alternative, leading to the selection of an underground storage, rather than storage in vessels. Also for Case 2c, the required storage volume is about 100,000 m3, while twice of this volume is required for Case 2f; it is noted that for these cases an additional back-up volume of about 6,400 m3 and 17,900 m3 of liquid nitrogen respectively for Case 2c and 2f is required in the ASU, due to the lower size of this unit and to allow base load operation of the gas turbines.

Hydrogen storage is not a novel industrial application. In fact, over the last decades there have been several examples of underground storage, like:

  • England, Teesside, Yorkshire: ICI has stored 1 million Nm3 of nearly pure hydrogen in three salt caverns at about 400 m in depth. The caverns have operated successfully for many years, and they are now operated by SABIC.
  • France, Beynes, Ile de France: Gaz de France has stored a gas with 50-60% hydrogen in an aquifer of 330 million Nm3 capacity for nearly 20 years. No losses or safety problems have been recorded.
  • Germany: 62% H2 gas was stored in a salt cavern of 32000 m3 at 80-100 bar
  • Texas: Praxair is constructing a large underground hydrogen storage facility in salt caverns, to enable “peak shaving” of its hydrogen production.

Figure 3.4-1: Case 2b – Balance of syngas within the week

Image

Table 3.4-1 summarizes the main performance and cost data of the three cases. From the figures in the table it can be drawn the following:

  • By introducing hydrogen storage in the plant, the electricity production during peak demand period increases by about 3% and 6% with respect to the reference case, respectively if the plant is required to generate the 50% of the net power output or to operate in island mode during low electricity demand period. In addition, the introduction of a PSA unit can allow to produce a significant amount of high purity hydrogen (75,400 Nm3/h).
  • For the hydrogen co-production alternative, the investment cost increase is about 3% of the reference case, while for the hydrogen storage case, the investment cost reduction is about 6% and 12.5%, respectively for Case 2c and 2f. These cost figures do not include cost for hydrogen storage, which depends both on the storage type (natural reservoir or mined cavern) and whether it is constant-pressure or variable-pressure storage. From literature data, it can be derived that the expected cost for the hydrogen storage of these IGCCs plant may vary from 10 M€ to 50 M€ (twice for Case 2f), corresponding to a maximum of 3% (6%) of the overall plant cost.

Table 3.4-1: H2 storage in IGCC plants – Performance and cost data summary (Est. accuracy: ±35%)

Tag Plant type Reference plant H2 storage in IGCC plants
Performance pipe diam. (mm) TIC, M€ Performance (peak time) pipe diam. (mm) Main changes TIC, M€
Case 2b IGCC w pre-comb capture NPO=730MWe 1,885 NPO=750MWe H2 storage working volume: 100,000 m3 1,931(w/o storage
H2 prod.: 75,400 Nm3/h
Case 2c IGCC w pre-comb capture NPO=730MWePipeline D: 500 1,885100 kmpipe: 206 NPO=754MWePipeline D: 450 PU % size of ref. plant: 82% 1,781(w/o storage)100 km pipe: 195
H2 storage working volume: 100,000 m3
Case 2f IGCC w pre-comb capture NPO=730MWePipeline D: 500 1,885100 kmpipe: 206 NPO=774 MWePipeline D: 450 PU % size of ref. plant: 65% 1,651(w/o storage)100 km pipe: 195
H2 storage working volume: 200,000 m3

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost; PU: Process Units

3.5 Oxygen storage in IGGC and oxy-USCPC power plants with CCS

The ASU significantly impacts the overall net electricity production of the plant, mainly due to its high auxiliary power demand. By reducing the energy requirement of this unit, at least during peak-demand hours, it is possible to increase the overall net power export during remunerative hours and improve the economics of the plant.

3.5.1 Oxygen storage for plants with two operating regimes

Two different design alternatives can be considered for either the IGCC or the oxy-combustion USCPC plant (Case 2a and Case 4b), both requiring adequate oxygen storage (as well as nitrogen storage for the IGCC), sized to cover production fluctuations of a cyclic operation, based on the electricity demand curve shown in section 3. The two scenarios assessed are the following:

  • Scenario 1 (partial load): ASU is operated at partial load during peak hours, while the rest of the plant runs at full load, thus reducing the auxiliary consumption and increasing the overall net electricity production.
  • Scenario 2 (reduced capacity): ASU is designed for a reduced capacity, with a consequent lower investment cost, while the plant load is changing in response to the variable electricity market requirements.

In both scenarios, oxygen from/to the storage system will need to be balanced during the weekly cyclic operation, in order to avoid any accumulation of the product. The need of balancing oxygen to/from the storage determines a relation between the ASU, running at low load during high electricity demand hours, and the other units, running at partial load during low electricity demand period. In fact, during off-peak operation the plant auxiliary demand and the resulting plant load strongly depend on the ASU load, which will need to ensure as a minimum the oxygen required by the plant to produce 50% of the daily power output, plus the oxygen sent to storage, necessary to fulfil the peak-hours demand.

For the oxy-combustion USCPC plant, during peak demand period compressed air is liquefied to provide the heat required for liquid oxygen from storage vaporisation. Liquid air is stored in pressurised vessel and vaporised during off-peak operation to replace the liquid oxygen sent to storage, in the main ASU exchanger.

Figure 3.5-1 shows the volume of stored oxygen during the week, for the two scenarios of Case 2a (similar trend is for Case 4b). The required net volume of the storage tank is the difference between the maximum and the minimum volume of stored oxygen during the week. From the graph, it can be concluded that it corresponds to the oxygen stored during the weekend, from the turndown of Friday night to the ramp up of Monday morning. A minimum oxygen storage volume corresponding to normal requirement of the plant, similarly to the reference plant, has been also considered while defining the tank size.

For the oxy-combustion USCPC plant, it is noted that oxygen storage has also been assessed in Case 4a of the study, in relation to the ramp rate of the Air Separation Unit, which is generally different, lower, than the one of a conventional boiler (typically 3% per min for vs. 4-5% per min for the PC boiler). In fact, by introducing a properly designed oxygen storage and vaporization system, it is possible not to affect the normal ramp-rate capacity of the boiler plant. The analysis showed that the difference between the ASU supply rate and the demand of the boiler is less than 10 tonnes of oxygen for each ramp-up phase. Therefore, the 200 tonnes back-up LOX storage tank and vaporiser system, already included in the reference design case, are also adequate to meet this requirement.

Table 3.5-1 and Table 3.5-2 summarize the main performance and cost data of the two power plants.

Figure 3.5-1: Case 2a –Stored Oxygen volume during the week

Image

Table 3.5-1: O2 storage (Scenario 1) – Performance and cost data summary (Estimate accuracy: ±35%)

Tag Plant type Reference plant Scenario 1 (ASU at partial load operation)
Performance TIC, M€ Performance (peak time) Main changes TIC, M€
Case 2a IGCC w pre-comb capture NPO=730MWeNEE=31.4% 1,885 NPO=786MWeNEE=33.9% O2 storage1 × 6,500 m3(D: 27.4 m; H: 11 m) 1,937
N2 storage1 × 17,500 m3(D: 43 m; H: 12 m)
New MACs.: 4 × 16 MWe
Case 4b Oxy-combustion USC PC w flue gas cryogenic purification NPO=533MWeNEE=35.5% 1,387 NPO=561MWeNEE=37.4% O2 storage1 × 10,500 m3(D: 33.5.1 m, H: 12.2 m) 1,422
Liquid air vessel4 × 1,600 m3(D: 8.8 m, H: 26.4 m)
2×60% Air compressors Booster compressor1 × 1.4 MWe

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost; MAC: Main air compressor

Table 3.5-2: O2 storage (Scenario 2) – Performance and cost data summary (Estimate accuracy: ±35%)

Tag Plant type Reference plant Scenario 2 (reduced ASU capacity)
Performance TIC, M€ Performance (peak time) Main changes TIC, M€
Case 2a IGCC w pre-comb capture NPO=730MWeNEE=31.4% 1,885 NPO=759MWeNEE=32.7% O2 storage:1 × 4,200 m3(D: 20.4 m, H: 12.8 m) 1,890
N2 storage1 × 6,500 m3(D: 27.4 m; H: 11 m)
ASU size: 82.5% of reference plant
New MACs.: 2 × 21 MWe
Case 4b Oxy-comb.USCPC w flue gas cryogenic purification NPO=533MWeNEE=35.5% 1.387 NPO=547MWeNEE=36.4% O2 storage 1 × 5,500 m3(D: 23.8 m, H: 12.8 m) 1,361
Liquid air vessel2 × 1,680 m3(D:9 m, H: 27 m)
ASU size: 78% of reference plant Booster compressor 1 × 0.75 MWe

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost; MAC: Main air compressor

From the figures in the tables the following conclusions can be drawn:

  • By introducing adequate oxygen (and nitrogen) storage in the plant and running the ASU at partial load, the electricity production during peak demand is about 5% and 8% higher than the reference case, respectively for Oxy-combustion and IGCC plant.
  • For the IGCC plant, the investment cost delta is about 3% higher than the reference case by considering an ASU at partial load operation, while it is approximately 2.5% for the Oxy-combustion plant.
  • By considering a lower-sized ASU (about 80% of the reference case), the electricity production is 3% and 4% higher than the reference case, respectively for the Oxy-combustion and the IGCC plant. Moreover, the total investment cost is about the same as the reference case for the IGCC, while for the Oxy-combustion power plant it is approximately 2% lower.

3.5.2 Oxygen storage for plants with three operating regimes

Cases 2g (IGCC plant) and 4d (Oxy-fuel plant) are based on a weekly demand curve characterized by the following three operating regimes:

  • Peak electricity demand period: 2 hours per working day.
  • Normal operation: 14 hours per working day.
  • Off-peak electricity demand period (50% of net power output): night and weekend.

During normal and peak electricity demand the IGCC is operated at base load to maximise the electricity production, while during off-peak electricity demand, the plant is required to produce 50% of the overall net electricity production capacity.

For the two hours of peak electricity demand, the ASU is operated at its minimum load and oxygen from the ASU is integrated with the oxygen coming from the liquid storages, after vaporisation. The minimum load is represented by the minimum technical load of the ASU cold box. i.e. around 50% of the design capacity. For the IGCC case, the air required by the ASU to obtain the 50% oxygen production is derived from gas turbine compressors while, for the oxy-combustion plant, a dual train configuration has been considered for the main air compressor to avoid inefficient operation at a load lower than 70%.

The oxygen requirement during peak hours is balanced by the production during night time, following a daily cycle operation and avoiding any accumulation of the stored product, thus implying a lower storage tank volume with respect to the weekly storage cycle scenarios.

Table 3.5-3 summarizes the main performance and cost data of the two power plants. From the figures in the tables the following conclusions can be drawn:

  • By introducing adequate oxygen (and nitrogen) storage in the plant and running the ASU at partial load, the electricity production during peak demand is about 6% and 10% higher than the reference case, respectively for the Oxy-combustion and the IGCC plant.
  • For the IGCC plant, the investment cost delta is about 1.5% higher than the reference case, while it is approximately the same for the Oxy-combustion plant.

Table 3.5-3: O2 storage (daily cycle) – Performance and cost data summary

Tag Plant type Reference plant Daily cycle LOX storage with an alternate demand curve
Performance TIC, M€ Performance (peak time) Main changes TIC, M€
Case 2g IGCC w pre-comb capture NPO=730MWeNEE=31.4% 1,885 NPO=806MWeNEE=34.7% O2 storage1 × 2,000 m3(D: 15.2 m; H: 11 m) 1,910
N2 storage1 × 1,450 m3(D: 13 m; H: 11 m)
New MACs.: 2 × 18 MWe
Case 4d Oxy-combustion USC PC w flue gas cryogenic purification NPO=533MWeNEE=35.5% 1,387 NPO=564MWeNEE=37.5% O2 storage1 × 600 m3(D: 9.1 m, H: 9.8 m) 1.399
Liquid air vessel1 × 230 m3(D: 4.8 m, H: 14.4 m)
2×50% Air compressors New air compressor 1 × 7MWe

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost; MACs: Main air compressors; Estimate accuracy: ±35%

3.6 Operation without carbon capture and sequestration

Provided that design is adequately made, power plants with CO2 pre or post-combustion capture can also be maintained in continuous operation without making the capture and compression of the carbon dioxide for transportation outside plant battery limits. Depending on possible low CO2 emission allowance costs, as in the present market situation, this operating flexibility may improve the economics of the plants, because of the resulting higher power production in this operating condition. However, a critical factor in determining whether a plant may be operated without capture is the acceptability of this approach to regulators.

Flexible CO2 capture operation is particularly suited for post-combustion CO2 capture systems (NGCC-Case 1e, USPC-Case3d), as it is possible to totally by-pass the CO2 capture unit, directly releasing to atmosphere the flue gases from the boiler, similarly to conventional power plants without CO2 capture. In this operating mode, the energy penalties related to the CO2 capture and compression units, as well as the steam requirement for solvent regeneration, are avoided, leading to an overall higher plant net power production. However, this implies that the whole cycle has to be designed for accepting all the steam from the steam generation, when the capture plant is turned off.

For IGCCs with pre-combustion CO2 capture processes (Case 2d), the Acid Gas Removal Unit cannot be shut down because it is necessary to remove at least the H2S from the syngas, before combustion in the Gas Turbine, to meet the design environmental emission limits. In addition, fuel composition to the gas turbine cannot be changed dramatically (e.g. CO shift unit cannot be by-passed) because it is necessary to respect the maximum range variation of fuel properties (e.g. LHV, Wobbe index etc.) as tolerated by the machine.

However, it is possible to tune to a certain extent the CO2 capture rate, and consequently the plant net power output, varying the solvent circulation flowrate in the AGR unit, in order to absorb completely the H2S but not the CO2. With this strategy, the capture rate range to which it is possible to operate is limited by both the AGR design and the gas turbine flexibility in accepting a variable fuel composition.

In the plant configuration assessed in Case 2d (IGCC), it has been considered that the AGR continues making the capture of the CO2 from the syngas: part of it is used as diluent in the gas turbine for NOx reduction and power augmentation, while the remainder is released to atmosphere, thus saving the CO2 compressor power demand. However, it is noted that the content of toxic components in the vented stream, in particular H2S and CO, does not allow its direct release to atmosphere. To overcome this problem, the following two alternatives have been considered:

  • Scenario 1: Different AGR unit design, to meet minimum H2S and CO specification for direct venting of the stream.
  • Scenario 2: Treatment and purification of the CO2 in a system downstream the AGR unit, without changing the design of the reference case.

For Scenario 1, with respect to the AGR design of the reference plant, major design changes of this configuration are the following:

  • Increased H2S absorber height and additional solvent chiller to meet the H2S specification in the CO2 vent stream.
  • Additional CO2 flash drum and recycle compressor to remove enough CO and meet CO2 vent stream specification.

As a consequence, these modifications lead to higher investment cost and higher steam and power consumptions of the unit, also when the plant is making full capture of the CO2 for delivery to plant battery limits.

For Scenario 2, the main drawback for venting the CO2 stream from the AGR is that the content of H2S in the stream is higher than 100 ppmv, while the benchmark limit value is assumed to be 5 ppmv. Several purification methods, based on sulphur absorption on catalyst bed, are proposed by specialised vendors, to meet the H2S specification in the venting stream. The main disadvantage of all these alternatives is the compression of the CO2 vent stream up to at least 20 bar, as required by the upstream purification treatment. In fact, lower pressure of the feed stream leads to excessive volumes of the reactors, and, consequently, of the catalyst required for the purification treatment. To reduce also the CO and H2 content in the CO2 vent stream, an additional treatment is required, based on the catalytic oxidation of these components. As for the H2S removal, the required amount of oxygen does not affect the ASU capacity. However, catalyst required for this purification treatment, typically based on platinum, can be poisoned by sulphur components.

The following table summarizes the main performance and cost data of the different power plants.

Table 3.6-1: Operation without CCS – Performance and cost data summary

Tag Plant type Reference plant Flexible plant operation
Performance TIC, M€ Performance Design modification TIC, M€
Case 1e NGCC w post-comb NPO=742MWeNEE=50.6% 726 Without CCSNPO=860MWeNEE=58.6% Greater ST LP module and condenser 768
With CCSNPO=736MWeNEE=50.2%
Case 2d IGCC w pre-comb: Scenario 1: modified AGR design NPO=730MWeNEE=31.4% 1,885 Without CCSNPO=777MWeNEE=33.5% Taller H2S absorber, additional chiller 1,895
With CCSNPO=722MWeNEE=31.1%
Case 2d IGCC w pre-comb: Scenario 2: treatment of CO2 vent stream NPO=730MWeNEE=31.4% 1,885 Without CCSNPO=747MWeNEE=32.2% Absorption catalyst bed 1,909
With CCSNPO=730MWeNEE=31.4%
Case 3d USCPC w post-comb NPO=666MWeNEE=34.8% 1,513 Without CCSNPO=848MWeNEE=44.3% Greater ST LP module and condenser; Condensate preheating line; Additional SW pumps 1,572
With CCSNPO=655MWeNEE=34.2%

Legend: NEE=Net Electrical Efficiency; NPO=Net Power Output; TIC=Total Investment Cost Estimate accuracy: ±35%

From the figures in the table the following conclusions can be drawn:

  • For the two post-combustion cases, the plant performances are same as the conventional plants without capture, but this option slightly reduces the efficiency and increases the costs when operating the plant with CCS. This is not the case for the IGCC, due to the maximum range variation of fuel properties tolerated by the gas turbine.
  • For the IGCC case, by considering an AGR design that meets minimum H2S and CO specifications for direct release of the CO2, the power production is 4% higher than the alternative with treatment and purification of the stream, while the investment cost is only marginally affected. However, a performance penalty shall be considered in normal operation with CO2 capture, the power production being 8 MWe less than the reference case.
  • By introducing a CO2 purification unit in the IGCC plant, the performances of the reference case are not affected, but approximately 30 MWe power production are lost while releasing the CO2 to atmosphere with respect to a modified AGR design. The total investment cost increase of the plant is about 1.3% higher than the reference case.