Carbon dioxide recovery from industrial processes

Energy Convers. Mgmr Vol. 36, No. 6-9, pp. 827-830, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0196-8904(95)00131-X 0196-8904/95 $9.50 + 0.00

CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES

JACCO CM. FARLA, CHRIS A. HENDRIKS, KORNELIS BLOK Department of Science, Technology and Society (STS), Utrecht University Padualaan 14, NL-3584 CH Utrecht, The Netherlands

Abstract - In this study possible techniques of recovering CO, from large-scale industrial processes are assessed. The largest and most concentrated CO, sources in Dutch industry are identified, and the technical and economic feasibility of recovering CO, from these sources is assessed. Nearly 20% of the Dutch industrial CO, emissions may be avoided. The mitigation costs are calculated to be between US$ 8 and US$ 46 per tonne of CO, emission avoided, for the different sectors considered. These cost figures indicate that CO, recovery from industrial processes and from power plants are competitive options. 1. INTRODUCTION To date, little attention has been paid to CO, recovery from industrial processes, although large amounts of CO, are emitted at high concentration by a few industries. It might be possible to recover carbon dioxide from these sources at a lower cost than from power plants. The objective of this study is to make a preliminary assessment of the possibilities of carbon dioxide recovery from large-scale industrial production processes. In this paper the largest and/or richest sources of CO, in industry are identified and the technical and economic feasibility of recovering CO, from these sources is assessed. We focus our attention on industries in the Netherlands. This paper is based on a more extended study by our department [I]. Specifications were set for the recovered CO, so that different CO, recovery options can be compared [2]. The product CO, should be delivered at a pressure of 110 bar, a temperature of 10°C and a water content of less than 10 ppm. Carbon dioxide with these specifications is regarded to be suitable for pipeline transport and subsequent underground storage. Costs are reported in US Dollars of 1990 (US$). Depreciation is calculated on the basis of annuity, with a depreciation time of 25 years and a real interest rate of 5%. The price of electricity is taken to be 0.05 US$/kWh. We use a steam price of 2.5 US$/GJ for low pressure steam (3.5 bar saturated). Energy used for the recovery of carbon dioxide leads to new carbon dioxide emissions, direct and indirect. To get a clear idea of the volume of CO, emissions avoided, we assigned carbon dioxide emission factors to the electricity and steam consumption. For electricity we use a carbon dioxide emission factor of 177 kg-CO,/GJ,; this figure is based on the fuel input in the Dutch public electricity production in 1988 [3]. For steam the carbon dioxide emission factor is taken as 62 kg-C02/GJ. This figure is based on fueling with natural gas with a thermal boiler efficiency of 90% (LHV). 2. CO, EMISSIONS IN THE NETHERLANDS The sources in industry from which CO, (or other carbon compounds) can be recovered are flue and fuel gases and feedstock gas streams. The carbon in these gases may be derived from fossil fuels, but in industry also carbonaceous gases are found that are not derived from fossil fuels (e.g. in the cement industry). In order to evaluate the possible ways for recovering CO,, all the CO, sources have to be mapped accurately. The actual CO, emission in the Netherlands is calculated to be 158 Mtonne CO, (1988, excluding international bunkers and import) [4]. Of this amount 50 Mtonne CO, was emitted by the manufacturing industry (excluding feedstock use of fuel). To obtain insight in the plants with the largest CO, emissions, we looked at the lowest level of aggregation, namely per plant. Estimates of the combined combustion and process CO, emissions resulted in a list of the top 20 Dutch C02-producing industrial plants. These 20 plants are grouped per industrial sector in figure 1. 827

828

FARLA et al.: CO? RECOVERY FROM INDUSTRIAL PROCESSES

3. INDUSTRIAL CO, RECOVERY From figure 1 it is clear that the largest Mt CO2 CO, emitting plants in the Netherlands are to be found among the refineries, in the basic metal industry, and in the petrochema ical and fertilizer industries. It is worth noting here that the heavy industries are 6 very important for the Dutch economy. Some heavy industry sectors (iron and steel, cement) are made up of only a few very large plants. The carbon dioxide emission from refineries has been described by Bakker [5]. In refineries gas streams with a high CO, concentration can be identified in hydrogen manufacturing units and in residue gasification plants. Bakker calculated that for a residue gasification plant with a CO, Fig. 1. The 20 industrial plants in the Netherlands with the production of 2100 tonne per day, the CO, highest COz-production in 1986. Each rectangle represents can be recovered at a cost of US$ 13 per one industrial site. tonne of CO, avoided [5]. However, this is only possible when these (planned) units are introduced in the Dutch refineries. In the following subsections, we deal with carbon dioxide removal in the fertilizer industry, the iron and steel industry and the petrochemical industry. 3.1. The Fertilizer Industrv In the Netherlands, the production of ammonia is responsible for over 90% of the energy consumption in the fertilizer industry. Ammonia is produced by steam reforming of natural gas. While ammonia does not contain any carbon, all the carbon has to be recovered during the production process. Approximately 1.2 tonne of carbon dioxide is recovered during the production of 1 tonne of ammonia. With an ammonia production of 3.65 Mtonne (1988), the amount of recovered CO, is 4.2 Mtonne-CO* per year. Part of the recovered CO, is used, the rest is vented to the atmosphere. Approximately 0.6 Mtonne-CO2 is used in the synthesis of urea [6]. Another 0.6 Mtonne-CO, is sold to CO, consumers (mainly the food and beverages industry). Based on these estimates of recovery and use, the amount of CO, that is recovered and subsequently vented to the atmosphere will be approximately 3 Mtonne-CO2 per year. Instead of venting, the carbon dioxide may be compressed in a four-stage isentropic compression process. The compression energy amounts to 393 kJ/kg-CO, (derived from: [71). Most of the water will be removed during the first compression stages. Additional drying consumes 8 kJ/kg-CO, of heat, and cooling takes 8 kJ/kg-CO, of electricity. 3.2. The Iron and Steel Industrv In the Netherlands, primary steel is produced at one iron and steel works only. Production involves reduction of iron-ore to pig iron in a blast furnace, and conversion to crude steel in a basic oxygen furnace (BOF). Approximately 5.3 Mtonne of crude steel was produced in 1986. The associated production of carbon-containing gases is estimated to amount to 8 Mtonne-CO*. During the production process three secondary gas flows evolve; coke-oven (CO) gas, blast furnace (BF) gas and basic oxygen furnace (BOF) gas. These three gas flows contain 85% of the carbon introduced into the process: approximately 70% in the BF gas, 9% in the CO gas and 7% in the BOF gas. The balance is incorporated in the steel, slag, and by-products. Since approximately 70% of the total carbon input emerges in the blast furnace gas, either as CO or CO,, recovery of the carbon dioxide from this gas will be described. The BF gas contains 20% (v/v) CO, [8]. At present, pig iron is produced in two blast furnaces with capacities of 6CKKland 8500 tonne of pig iron per day. The BF gas evolves from the top of the blast furnace. After dust removal the gas is expanded for power recovering in turbines [9]. Activated MDEA (methyldiethanolamine) is a suitable solvent for the separation of CO, from blast furnace gas [lo]. The chemical absorption unit is placed between the dust cleaning step and the expansion turbine. Apart from the dust removal there will be no need for further clean-up of the BF gas prior to CO, recovery. The HZS that is recovered together with the CO,, will give approximately 0.5% H,S in the product gas. For the recovery of carbon dioxide from BF gas, the most economic plant design is a one-stage chemical absorption configuration [lo]. The heat consumption is

FARLA et al.: CO2 RECOVERY FROM INDUSTRIAL PROCESSES

829

estimated to be 140-160 kJ/mole CO, [lo]. The electric energy consumption is estimated to be 29 kJ/kg of CO, (derived from: [1 I]). After recovery, the compression energy will amount to 273 M/kg CO, (based on data from: 171). Drying of the CO, will consume 8 kJ, per kg CO,. Cooling of the CO, will take 8 kJe/kgCO,. Due to the recovery of CO,, less power will be generated in the expansion turbines. If 90% of the CO, is recovered, the electric power generation will fall from approximately 19 MW to 15 MW, representing a loss of electricity production equivalent to 43 kJ per kg of CO, recovered. 3.3. The Petrochemical Industry The three largest petrochemical industries in the Netherlands, depicted in figure 1, are estimated to emit 7.3 Mtonne of CO, annually (including process emissions). The production of ethylene, propylene, butadiene and benzene accounted for over 80% of the total fuel consumption in the petrochemical industry [12]. The CO, emissions associated with these products are combustion emissions. We estimate that it is possible to recover the CO, from 90% of the stack gases. The CO, from the stack gases can be recovered by means of a chemical absorption process. Because of the low CO, partial pressure (- 0.1 bar), monoethanolamine (MEA) is chosen for the absorption process. The heat consumption of the absorption process is estimated to be 4.2 MJ/kg-CO, [13]. The electricity consumption is estimated to be 29 kJ/kg-CO,. The flue gases are compressed to compensate for the pressure drop in the absorption column. The electric energy consumption for the stack gas blowers is calculated to be 72 kJ/kg of CO, (based on a CO, concentration of 10% in the flue gases). After recovery, the carbon dioxide is compressed and dried for transport. The compression energy will be 370 kJ/kg of CO, (derived from data by Smit [7]). Drying of the CO, is estimated to consume 8 kJ/kg-CO,. The total electric energy for cooling is estimated to be 8 W/kg of CO,. 4. ECONOMIC EVALUATION A preliminary cost estimate was made for CO, recovery in the three branches of industry described in the previous section. The cost estimate includes delivery of the CO, according to the set specifications. The recovery processes were designed to meet the capacity requirements of specific industrial plants in the Netherlands. Investment costs were obtained from literature and manufacturers of the equipment [7, 13-151. An installation factor of 2 was applied to the investment costs. When necessary, investment costs were converted with a scaling power factor of 0.7 [16]. The investment costs for the refrigeration step are so small that they are neglected in this evaluation. The cooling water supply is also neglected in the economic evaluation. O&M costs are taken to be 2.1% and 3.6% of the investment costs, for static and rotating equipment respectively. Other starting-points for the economic evaluation are described in the introduction. The cost estimates are reported in Table 1. In addition, we identified one small high concentration COz-source in the petrochemical industry, related to the ethylene oxide production. It is estimated that over 0.2 Mtonne of carbon dioxide emission can be avoided per year at a cost of 9 US$ per tonne of CO, avoided [l]. Table 1. Cost estimates for the removal of CO,.

Plant capacity

(Mt-CO,/yr)

Fertilizer Industry

Iron and Steel Industry

Petrochemical Industry

0.7

2.8

1.8

Electricity consumption

(GJ/t-CO*)

0.40

0.3sa

0.50

Steam consumption

(GJ/t-CO,)

0.01

3.40

4.20

Investment costs

(106 US$)

10

375

250

Total annual costs

(106 US$)

5

70

55

(Mt-CO,/yr)

0.7

2.0

1.2

(US$/t-CO, avoided)

8

35

46

CO, avoided Specific mitigation costs a Electricity consumption

includes the loss of electricity production in the blast furnace expansion turbines.

5. DISCUSSION AND CONCLUSIONS In figure 2 the results are given as a supply-curve. From this figure we may conclude that nearly 20% of the CO, emissions from the Dutch manufacturing industry may be avoided at a cost of between 8 and 46 US$ per tonne. These cost figures can be compared with cost figures for CO, recovery from power production plants (See also Fig. 2). The lowest cost figure reported by Hendriks [17] is between 14 and 17 US$ per

830

FARLA et al.: CO? RECOVERY FROM INDUSTRIAL PROCESSES

of CO, avoided, for an ICGCC with shift conversion of the fuel gas. This cost figure is considerably higher than the Petrochemical Industry cost of recovery and compression of CO, from the ammonia and ethylene oxide production. Recovery of CO, from the Iron and steel flue gases of power plants is reported to Industry cost US$ 3540 [17]. These figures are comparable with the cost figures for the recovery of CO, in the iron and steel industry, but lower than the estimated Refineries I residue gasification cost of recovery of CO, from the stack Ethylene oxide production gases in the petrochemical industry. Thus, as far as costs are concerned, CO, recovery from industry and from power plants % 1 2 3 4 5 6 7 6 9 CumulativeCO2 emission avoided (Mton) are competitive options. Therefore, if CO, recovery is to be applied, recovery Fig.2. Supply-curve for CO, recovery in Dutch manufacturing from industrial processes should certainly industry. be taken into consideration. One should be aware of the fact that layout and operational considerations (e.g. start up problems) were not part of this study. Furthermore, it is important to realize that current or future changes in production processes and product mixes will influence the CO, recovery possibilities described in this study. tonne

ACKNOWLEDGEMENT This study was executed as part of the “Integrated Research Programme on Carbon Dioxide Removal and Storage”. This programme was supported financially by the Dutch Ministry of the Environment, and by the Dutch National Research Programme on Global Air Pollution and Climate Change. The authors like to thank Ernst Worrell for his valuable comments on an earlier draft of this paper.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13. 14. 15. 16. 17.

REFERENCES Farla, J.C.M., CA. Hendriks, and K. Blok, Carbon Dioxide Recovery from Industrial Processes, Report 92076, Utrecht University, Dept. of STS, Utrecht, 1992. Blok, K., Final Report of the Integrated Research Programme on Carbon Dioxide Recovery and Storage, Report 92063, Utrecht University, Dept. of STS, Utrecht, 1993. Heijningen, R. van (ed.), Reductie van de Kooldioxide-uitstoot via het Afvalstoffenbeleid, Report 140346.01, Ministry of Housing, Physical Planning and Environment VROM, The Hague, 1990. RIVM (eds.) Nationale Milieuverkenning 2; Rijksinstituut voor Volksgezondheid en Milieuhygiene, Samsom H.D. Tjeenk Willink bv., Alphen a/d Rijn, 1991. Bakker, H., CO, Emission from Refineries, Report 60906-01, Comprimo, Amsterdam, 1992. Worrell, E., Personal communication, Utrecht University, Dept. of STS, Utrecht, 1992. Smit, J.B., Written communication, Sulzer Nederland b.v., Zoetermeer, 1992. IISI, Energy and the Steel Industry, International Iron and Steel Institute, Brussels, Belgium, 1982. Wolthuizen, S., IJzer, Staal en Walstechnologie, Hoogovens, IJmuiden, 1988. Vannby, R., Written communication, reJ: 201OORV/NM, Haldor Tops0 A/S, Lyngby, Denmark, 1992. Gerhardt, W. and W. Hefner, BASF’s Activated MDEA - A Flexible Process to Meet Specific Plant Conditions, in AIChE Ammonia Safety Symposium, Denver, CO, USA, 1988. Worrell, E, J.G. de Beer, R.F.A. Cuelenaere, and K. Blok, ICARUS, The Potential for Energy Conservation in the Netherlands up to the Year 2000, Report 92024, Utrecht University, Dept. of STS, Utrecht, 1992. Mariz, C., Written communication, Fluor Daniel, Inc., Irvine, CA, USA, 1991. Smit, J.B., Personal communication, Sulzer Nederland b.v., Zoetermeer, 1992. KTI, Removal of CO, from Reformer Gas in a Power Plant, Report vrom 262.848, Kinetics Technology International, Zoetermeer, 1992. Remer, D.S. and L.H. Chai, Design Cost Factors for Scaling-up Engineering Equipment, Chemical Engineering Progress, August 1990, p. 77-82. Hendriks, C.A., Carbon Dioxide Removal from Coal-Fired Power Plants, Kluwer Academic Publishers, Dordrecht, 1994.