Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu

Framing the cost analysis for CO₂ sequestration

The costs¹¹ of alternative options for hydrogen production and use/ CO₂ disposal are estimated here for hydrogen produced from natural gas, coal, and biomass, with a focus on automotive applications of the produced hydrogen.¹² Storage of CO₂ in both depleted natural gas fields and saline aquifers is considered. Various hydrogen production/ CO₂ disposal scenarios are investigated. Costs are estimated for hydrogen production, for delivering hydrogen to users, and for using hydrogen. The estimated costs for CO₂ sequestration include the costs for drying and compressing the CO₂ to the pressures required for CO₂ transport and injection,¹³ the costs of pipelines for transporting the CO₂ to the sequestering sites,¹⁴ and the costs for wells and surface facilities at the storage sites.¹⁵

Hydrogen production cost estimates under alternative assumptions about sequestration are presented in table 6.5 for natural gas, table 6.6 for coal, and table 6.7 for biomass feedstock’s, as functions of feedstock prices. In all cases it is assumed that the CO₂ separated (at 1.3 bar) at the hydrogen production facility must be dried and compressed to 110 bar for delivery to a CO₂ pipeline or natural gas disposal well. In table 6.8 hydrogen costs to consumers for transport applications are presented for base case scenarios for each feedstock as a function of the feedstock price, without and with sequestration, assuming typical sequestration costs in the latter instance.

For hydrogen production from natural gas with sequestration, it is assumed that the hydrogen plant is sited at a depleted natural gas field, in which the recovered CO₂ is sequestered, thereby avoiding long-distance CO₂ transport costs. Credit is taken for the modest increase in natural gas production that results from depressurization of the natural gas reservoir. It is assumed that: (i) as in the analysis of Blok et al. (1997), the increased natural gas production would contribute between 3.5 and 8.5 per cent of the natural gas feedstock requirements for the hydrogen production plant; (ii) increased production takes place after 20 years of primary production; (iii) after 15 years of CO₂ injection there is no more increased production; and (iv) the benefits of increased production accrue to producers of hydrogen from natural gas for the situation where conversion plants are located near the disposal site.

Because natural gas field disposal sites are geographically limited, it is assumed for the base case scenarios that in order to reach typical final consumers the produced hydrogen is transported 1,000 km further than is required for hydrogen that might be produced from coal or biomass at sites other than at depleted natural gas fields. To take advantage of the scale economies of pipeline transport of hydrogen, it is assumed for the base case scenarios that the outputs of five large production plants (each having a production capacity of 19.1 PJ of hydrogen per year - see table 6.5) are combined for hydrogen transport in a single 1,000 km pipeline. The cost of hydrogen to consumers is estimated by assuming that, after the initial 1,000 km hydrogen transport journey, the produced hydrogen is further transported 100 km in intermediate-sized pipelines, followed by transport 10 km in small diameter pipelines to refueling stations for automobiles, where the final costs tabulated in table 6.9 are incurred.

Costs for hydrogen from natural gas with sequestration are compared with costs for two alternative configurations for making hydrogen from natural gas without sequestration: one in which hydrogen is produced near the natural gas field and piped to distant markets; and another in which natural gas is piped to a hydrogen production facility near these distant markets (see table 6.8). Siting the facility for producing hydrogen from natural gas near hydrogen markets for one of the cases without sequestration is included because pipeline transport is less costly for natural gas than for hydrogen.

Table 6.5 Cost of producing hydrogen from natural gas near depleted natural gas fields (US$/GJ of H₂)

a. Lifecycle emissions are 21.8 kg C/GJ of H₂ with venting of the separated CO₂, and 11.1 kg C/GJ if the separated CO₂ is sequestered (see table 6.1).
b. For a plant producing 19.1 PJ/year of H₂ at 75 bar via steam re-forming. The overnight construction cost is US$188.3 million; land + working capital are US$18.2 million; external electricity requirements are 8.2 kWh/GJ of H₂ (Williams et al. 1995a,b). Assuming a 3-year construction period, a 25 year plant life, and an annual insurance charge of 0.5 per cent, the annual fixed capital charge rate is 0.1277, while the annual capital charge rate for land + working capital is 0.10. It is assumed that the purchased electricity price is US$0.05/kWh. Here P_ng is the price of natural gas (NG) feedstock, in US$/GJ.
c. Following Blok et al. (1997), it is assumed that the extra NG available from depressurizing the reservoir via CO₂ injection is 3.5-8.5 per cent of total NG requirements averaged over a 15-year period after primary NG production ceases; the value of enhanced production levelized over the 25-year life of the plant is 2.9-7.1 per cent of NG feedstock needs.
d. CO₂ is produced at the H₂ plant in a 750,000 tonnes/year stream at 1.3 bar and 40°C, with 1.5 per cent moisture (Katofsky 1993). It is assumed that the CO₂ must be dried to less than 10 ppm and compressed to 110 bar (Farla et al. 1992). The capital cost CC = US$10.35*106US$) from CC = 0.351*(CAP)^0.51117 (Farla et al. 1992), where CAP is the capacity (in 10³ tonnes of CO₂/year). Assuming a 1-year construction period a 25 year plant life, and an insurance charge of 0.5 per cent per year, the annual capital charge rate is 0.1172. It is assumed that annual O&M costs are 3.5 per cent of the capital cost. In this capacity range, electricity requirements, EL (in kWh/tC), are given by EL = 1097.5*(CAP)^-01509 (Farla et al. 1992), or 404 kWh/tC.
e. For injection into depleted NG fields at a depth of 2 km, Hendriks (1994) has calculated that storage costs range from US$2.6 for an injection rate of 20 Nm³/s to US$13.3/tC for an injection rate of 2 Nm³/s.

Table 6.6 Cost of producing hydrogen from coal (US$/GJ of H₂)

a. Lifecycle emissions are 41.2 kg C/GJ of Hz with venting of the separated CO₂, and 11.4 kg C/GJ if the separated CO2 is sequestered (see table 6.1).
b. For a plant producing 37.8 PJ/year of H₂ at 75 bar using a Shell oxygen-blown gasifies, for which the overnight construction cost is USS924.4 million; land + working capital are US$78.2 million; external electricity requirements are 27.8 kWh/GJ of H₂ (Williams et al. 1995a,b). Assuming a 4-year construction period, a 25-year plant life, and an annual insurance charge of 0.5 per cent, the annual fixed capital charge rate is 0.1337, while the annual capital charge rate for land + working capital is 0.10. It is assumed that the purchased electricity price is US$0.05/kWh. Here P_c is the price of the coal feedstock, in US$/GJ.
c. From Farla et al. (1992) the cost of drying the CO₂ to <10 ppm is US$0.073/tC for steam (at 8 MJ/tonne of CO₂) plus US$0.416/tC for capital (for a capital cost CC = 1.48*(CAP)0.7 = US$3.99 million when CAP = CO₂ recovery rate = 4.125*10⁶ tonnes/year of CO₂; assuming a 25-year plant life and an annual insurance charge of 0.5 per cent per year, the annual capital charge rate is 0.1172) plus an annual O&M cost of 2.1 per cent of the capital cost per year, or US$0.074/tC. Thus the total cost of drying is US$0.563/tC, or US$0.017/GJ of produced H₂.
d. The cost of transporting coal from where it is produced to a depleted NG field site for a hydrogen production facility is taken to be US$0.35/GJ, the difference between the import price of Australian coal (with a higher heating value of 27.2 GJ/tonne) imported into Europe and its export price from Australia in 1993 (IEA 1995). This translates into a cost penalty of US$0.55/GJ of produced hydrogen.
e. Skovholt (1993) has calculated the pipeline cost, PC, in US$/tC (including costs of compression of CO₂ from atmospheric pressure to 110 bar) for 250 km of pipeline transmission of CO₂, for pipelines of capacities of 3-100 million tonnes of CO₂/year. A regression yields PC= 43.85*(CAP)^-0.54786, where CAP is the capacity in million tonnes of CO₂/year. For a plant with CAP = 4.125 million tonnes of CO₂ per year, PC = US$20.18/tC or US$0.60/GJ of H₂.
f. For injection into saline aquifers (depleted NG fields) at a depth of 2 km, Hendriks (1994) has calculated the storage cost to range from US$9.2/tC (US$2.6/tC) for an injection rate of 20 Nm³/s to US$34.3/tC (US$13.3/tC) for an injection rate of 2 Nm³/s.

Table 6.7 Cost of producing hydrogen from biomass (US$IGJ of H₂)

a. Lifecycle emissions are 5.4 kg C/GJ of H2 with venting of the separated CO₂, and -18.4 kg C/GJ if the separated CO₂ is sequestered (see table 6.1).
b. For a plant producing 7.73 PJ/year of H₂ at 75 bar using a Battelle Columbus Laboratory indirectly heated biomass gasifies, for which the overnight construction cost is US$170.3 million; land + working capital are US$14.7 million; external electricity requirements are 21.7 kWh/GJ of H2 (Williams et al. 1995a,b). Assuming a 2-year construction period, a 25-year plant life, and an annual insurance charge of 0.5 per cent, the annual fixed capital charge rate is 0.1215, while the annual capital charge rate for land + working capital is 0.10. It is assumed that the purchased electricity price is US$0.05/kWh. Here Pb is the price of the biomass feedstock, in US$/GJ.
c. From Farla et al. (1992) the cost of drying the CO₂ to <10 ppm is US$0.073/tC for steam (at 8 MJ/tonne of CO₂) plus US$0.714/tC for capital (for a capital cost CC = 1.48*(CAP)^0.7 = US$1.12 million when CAP = CO₂ recovery rate = 0.674*10⁶ tonnes/year of CO₂; assuming a 25-year plant life and an annual insurance charge of 0.5 per cent per year, the annual capital charge rate is 0.1172) plus an annual O&M cost of 2.1 per cent of the capital cost per year, or US$0.128/tC. Thus the total cost of drying is US$0.915/tC, or US$0.022/GJ of produced H₂.
d. Assuming that wet biomass has a higher heating value of 10 GJ/tonne and that the cost per tonne of transporting biomass or coal from where it is produced to a hydrogen production facility at a depleted NG field site is the same, the transport cost per GJ for biomass would be 2.7 times the cost for Australian coal, or US$0.94/GJ if the benchmark for coal is as assumed in note d, table 6.6. This translates into a cost penalty of US$1.48/GJ of produced hydrogen.
e. Following the approach outlined in note e, table 6.6, the pipeline cost in US$/tC for 250 km of pipeline transmission of CO₂ is PC= 43.85*(CAP)^-0.54786 = US$54.44/tC or US$1.30/GJ of produced hydrogen, for CAP = 0.674 million tonnes of C0₂/year.
f. For injection into saline aquifers (depleted NG fields) at a depth of 2 km, Hendriks (1994) has calculated the storage cost to range from US$9.2/tC (US$2.6/tC) for an injection rate of 20 Nm³/s to US$34.3/tC (US$13.3/tC) for an injection rate of 2 Nm³/s.

Table 6.8 Delivered costs of hydrogen produced from alternative feedstocks^a (US$/GJ of H₂)

a. Assuming average values for sequestration costs in tables 6.5, 6.6, and 6.7. Here P_ng is the wellhead NG price (in US$/GJ); P_c and P_b are, respectively, the prices of coal and biomass delivered to the conversion facility.
b. For the pipeline transmission and distribution (T&D) system associated with producing H2 from NG near the NG field, it is assumed that the output for a cluster of five plants (a production rate of 95.45 PJ/year or 850.3 million suf/day) is transported 1,000 km in a 81.3 cm (32 inch) pipeline, for which the inlet and outlet pressures are 75 bar (1,087.5 psia) and 48 bar (697 psia), respectively. At the end of this pipeline, the output is divided into 5 equal parts, each of which is transported another 100 km in 30.5 cm (12 inch) pipelines, for which the outlet pressure is 20.3 bar (294 psia). At the ends of these pipelines the H₂ is transported 10 km further to refuelling stations in 4.2 cm (1.7 inch), I million scf/day (0.113 PJ/year) lines for which the outlet pressure is 14 bar (200 psia).
c. In this instance NG is piped from the NG field via 1,000 km and 100 km pipelines (similar to those described in note b) to a H₂ conversion plant at the city gate. The cost of compressing the NG and transmitting it to the city gate adds US$0.64/GJ to the cost of Hz production. Because the plant is so close to the market, the H₂ recovered from the PSA unit need not be compressed to high pressure at the production plant. Here it is assumed that H₂ is instead sold at the 20.3 bar pressure at which it is recovered from the PSA unit, saving US$0.35/GJ in compression costs. This compression saving is much larger than the cost of compressing NG at the gas field to 75 bar. The lifecycle CO₂ emissions are reduced slightly as a result of this compressor work savings.
d. It is assumed that the output for a single plant producing H₂ from coal (a production rate of 37.76 PJ/year or 336.5 million scf/day) is transported 100 km in a 32.3 cm (12.7 inch) pipeline, for which the inlet and outlet pressures are 75 bar and 20.3 bar, respectively. At the end of this pipeline, the H2 is transported 10 km further to refuelling stations in 4.2 cm (1.7 inch), 1 million scf/day lines for which the outlet pressure is 14 bar.
e. It is assumed that the output for a single plant producing H2 from biomass (a production rate of 7.73 PJ/year or 68.9 million scf/day) is transported 100 km in a 17.7 cm (7 inch) pipeline, for which the inlet and outlet pressures are 75 bar and 20.3 bar, respectively. At the end of this pipeline the H2 is transported 10 km further to refuelling stations in 4.2 cm (1.7 inch), I million scf/day lines for which the outlet pressure is 14 bar.
f. Hydrogen T&D costs presented here are illustrative rather than definitive. (The pipeline/compressor systems are not optimized.) Flow rate and cost calculations are based largely on Christodoulou (1984). However, it is assumed that pipeline unit costs are never lower than about US$180/m, the actual installed cost for a 2,700 m pipeline rated for 1,000 psia with D = 3 inches (Ogden et al. 1995); thus, for small-diameter pipelines, it is assumed that the cost per metre of pipe is independent of the pipe diameter. It is assumed that pipelines last 50 years, so that the annual capital charge rate (including an insurance rate of 0.5 per cent per year) is 0.1059.
g. See table 6.9.
h. This is the carbon tax (CT) required to make the cost of H₂ produced from NG with sequestration in the depleted NG field equal to the cost of H₂ produced from NG at the city Bate without sequestration. At this tax level, the cost of H₂ produced from coal with sequestration is also less than the cost of H₂ produced form coal without sequestration.

Two scenarios are considered for coal and two for biomass. In one scenario for each feedstock, the base case, it is assumed that the recovered CO₂ is transported by pipeline 250 km to a site where sequestration is feasible in a saline aquifer. Because saline aquifers are so widely available, it is assumed that the pipelines required to bring the produced hydrogen to market are much shorter than for the case where hydrogen is produced from natural gas with sequestration. Specifically, it is assumed that the produced hydrogen is transported 100 km in an intermediate-sized pipeline, followed by transport 10 km in small-diameter pipelines to refuelling stations for automobiles, where the final costs tabulated in table 6.9 are incurred. In the other scenario the coal or biomass is transported to a hydrogen production facility located near a depleted natural gas field, where the separated CO₂ is sequestered. However, it is assumed in this instance that: (i) natural gas field disposal of CO₂ occurs after disposal has already taken place there for hydrogen produced from natural gas, and (ii) unlike the natural gas case, no additional enhanced natural gas production can be credited against the cost of hydrogen produced from coal or biomass.

The costs of hydrogen to consumers from natural gas, coal, and biomass feedstocks, as functions of feedstock costs, are presented in table 6.8, without and with sequestration, and without and with a carbon tax of US$52/tC - a tax large enough to make the cost of hydrogen from natural gas with sequestration at the gas field equal to that for hydrogen produced from natural gas at a site 1,100 km from the natural gas field, near large remote hydrogen markets, when the wellhead natural gas price is US$3/GJ, and for average sequestration cost and average credit for enhanced natural gas recovery (see table 6.5). The charge against the fuel cost to the consumer from this carbon tax is levied at the net lifecycle CO₂ emission rate for a given fuel, as indicated in table 6.1. Hydrogen costs to consumers, along with gasoline costs for comparison, are also shown in figures 6.2 and 6.3 for specific feedstock prices - both without sequestration and for low and high estimates of sequestration costs. (All fuel taxes are excluded in these figures.) In figure 6.2, these costs are presented per unit of delivered energy (US$ per GJ); in figure 6.3, they are presensed per unit of transport energy services provided (cents per km of driving), assuming the hydrogen is used for automobiles. The estimated total lifecycle costs of owning and operating fuel cell cars fuelled by hydrogen derived from alternative sources are shown in figure 6.4.

Table 6.9 Estimated cost of hydrogen refueling^a

a. For a large refuelling station providing 10⁶ scy/day (0.132 million GJ/year) of gaseous hydrogen, based on a design by Ogden et al. (1995). It is assumed that fuel cell cars have a 250 mile (400 km) range and that the average fuel economy of these cars is 71.6 mpg (30.4 km/litre) of gasoline equivalent, so that refuelling requirements are 0.46 GJ/car, and 787 cars are refuelled per day. Further it is assumed that H2 is stored onboard cars at 552 bar (8,000 psia), and that the compressor discharge pressure at the refuelling station is 558 bar (8,400 psia).
b. Storage cylinders for the refuelling station having a maximum operating pressure of 8,400 psia, a capacity of 6,005 scf each, and an installed cost per vessel of US$10,500. The refuelling station needs 150 cylinders, so that the total capital cost is US$1.575 million. The cylinders are expected to last 10 years. The annual capital charge rate (including an insurance rate of 0.5 per cent per year) is thus 0.1677.
c. The compressor capacity required for the refuelling station (for inlet and outlet hydrogen pressures of 200 psia and 8,400 psia, respectively) is 270 kW_e, its installed cost is estimated to be USS1,919/kWe, and its lifetime is expected to be 100,000 hours (11.4 years), so that the annual capital charge rate (including an insurance rate of 0.5 per cent per year) is 0.1559.
d. Ogden et al. (1995) estimate the annual cost for the compressor to be US$3,000.
e. Electricity requirements are 6.49 kWh/1,000 scf (17.93 kWh/GJ) of H₂. It is assumed that the electricity price is US$0.06/kWh.
f. The capital costs of the hydrogen dispenser, priority panel, and sequencer are estimated to be US$285,500. This equipment is expected to last 10 years. The annual capital charge rate (including an insurance rate of 0.5 per cent per year) is thus 0.1677.
g. The annual labour cost is US$131,400 per year, assuming a labour plus benefits rate of US$15/hour.

The feedstock prices assumed for the construction of figures 6.2, 6.3, and 6.4 are indicative for the period near 2010 when hydrogen might begin to come into the market in some areas: US$3.0/GJ for natural gas, US$1.35/GJ for coal, and US$2.0/GJ for biomass. The gasoline cost shown in these figures is for reformulated gasoline derived from US$23/barrel (US$3.75/GJ) crude oil (the world oil price projected for 2010 by the US Department of Energy - EIA 1995). The assumed natural gas price is the average natural gas wellhead price projected for the year 2010 for the United States by the US Department of Energy (EIA 1995), while the coal price is the average price projected for US electric utilities in that year (EIA 1995). The biomass price is what could be widely realizable for plantation biomass in developing countries in this time-frame, based on commercial plantation technology from Brazil;¹⁶ for the industrialized countries, plantation biomass prices based on present technology would be higher. However, if R&D goals for plantation biomass can be realized, biomass prices as low as US$1.5/GJ appear to be feasible for the time-period near 2020 for large-scale plantation biomass production in the United States.¹⁷ It is not likely that biomass would begin to be used for hydrogen production before 2020.

Contents - Previous - Next