CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES JACCO C. M. FARLA, CHRIS A. HENDRIKS and KORNELIS BLOK Department of Science, Technology and Society, Utrecht University, Padualaan 14, NL-3584 CH Utrecht, The Netherlands

Abstract. The ongoing human-induced emission of carbon dioxide (CO2) threatens to change the earth's climate. One possible way of decreasing CO2 emissions is to apply CO2 removal, which involves recovering of carbon dioxide from energy conversion processes and storing it outside the atmosphere. Since the 1980's, the possibilities for recovering CO2 from thermal power plants received increasing attention. In this study possible techniques of recovering CO2 from large-scale industrial processes are assessed. In some industrial processes, e.g. ammonia production, CO2 is recovered from the process streams to prevent it from interfering with the production process. The CO2 thus recovered can easily be dehydrated and compressed, at low cost. In the iron and steel industry, carbon dioxide can be recovered from blast furnace gas. In the petrochemical industry CO2 can be recovered from flue gases, using low-temperature heat for the separation process. Carbon dioxide can be recovered from large-scale industrial processes and in some cases the cost of recovery is significantly less than CO2 recovery from thermal power plants. Therefore this option should be studied further and should be considered if carbon dioxide removal is introduced on a wide scale.

1. Introduction According to the Intergovernmental Panel on Climate Change, the ongoing emissions of greenhouse gases from human activities are leading to an enhanced greenhouse effect. This may result, on average, in additional warming of the earth's surface [1]. Carbon dioxide (CO2) is a major greenhouse gas. To prevent a possible climate change the emission of CO2 should be reduced. Many options are available to reduce CO2-emissions to the atmosphere. These include reducing the use of energy derived from fossil fuels. Another possibility is to switch from coal to natural gas. Afforestation can also limit the increase of the atmospheric CO2 concentration. Another option that has received much attention lately is carbon dioxide removal: recovery of carbon dioxide from energy conversion processes, followed by storage outside the atmosphere. It is not likely that one single option (e.g. renewables) can avert the threat of a climate change. Therefore, all the available options should be investigated with respect to their potential and costs. It is also important to find out how these options can be applied in an environmentally sound way. This paper is concerned with carbon dioxide removal. Generally, CO2 recovery costs will be lowest where the CO2 concentration in a gas stream is high. Therefore, the search for sources where CO2 can be recovered Climatic Change 29: 439-461, 1995. (~) 1995 Kluwer Academic Publishers. Printed in the Netherlands.

440

JACCO C. M. FARLA ET AL.

TABLE I Specifications for the recovered CO2 Pressure Temperature Water content Permanent gases

110 bar 10 ° C _< 10 ppm _< 5% (mole/mole)

at low cost, is really a search for concentrated CO2 gas streams. Furthermore, advantage of scale suggests that recovery will be cheaper if CO2 sources are big. These two criteria are met in thermal power plants and in heavy industry. The recovery of carbon dioxide from thermal power plants has been studied extensively, and many cost estimates have been made (e.g. [2]). The lowest cost figure reported by Hendriks et al. [3] is U.S.$ t7 per tonne of CO2 avoided for a newly built IGCC with shift conversion of the fuel gas. Recovery of CO2 from the flue gases of a thermal power plant, by means of chemical absorption, is reported to cost U.S.$35 and U.S.$40, for coal-fired and natural gas-fired power plants respectively [3]. To date, little attention has been paid to CO2 recovery from industrial processes, although large amounts of CO2 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 object 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 CO2 in industry are identified and the technical and economic feasibility of recovering CO2 from these sources is assessed. We initially focus our attention on industries in the Netherlands, subsequently we will evaluate the possibilities of CO2 recovery for a broader region, i.e. the European Union (E.U.). 1.1.

ASSUMPTIONS

Specifications were set for the recovered CO2 so that different CO2 recovery options can be compared. These specifications are summarized in Table I. Carbon Dioxide with these specifications is suitable for pipeline transport and subsequent underground storage. Costs are reported in U.S. Dollars of 1990 (U.S.$) [4]. Depreciation is calculated based on annuity, with a depreciation time of 25 years and a real interest rate of 5%. The price of electricity is taken to be 0.05 U.S.$/kWh. We use a steam price of 2.5 U.S.$/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 CO2 emissions

CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES

441

avoided, we assigned carbon dioxide emission factors to the electricity and steam consumption. For electricity we use a carbon dioxide emission factor of 177 k g CO2/GJe; this figure is based on the fuel input in the Dutch public electricity production in 1988 [5]. For steam the carbon dioxide emission factor is taken as 62 kg-CO2/GJ. This figure is based on fuelling with natural gas with a thermal efficiency of 90% (LHV) [6].

2. Large-Scale CO2 Emissions The worldwide carbon dioxide emission from fossil fuel combustion is estimated to be 22 + 2 Gigatonne CO2 in 1990 [7]. From this amount, the estimated CO2 emission from power plants is 5.9 Gtonne-CO2/yr [8]. The estimated CO2 emission from industry is 4.8 Gtonne-CO2/yr [8]. The largest part of these emissions arises from the heavy industry sectors. A further subdivision of worldwide CO2 emissions for the different industrial sectors is not available in the literature. The sources in industry from which CO2 (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 CO2 recovery possibilities, all the CO2 sources have to be mapped accurately. 2.1. C O 2 EMISSIONS IN THE NETHERLANDS

The carbon dioxide emission in the Netherlands is calculated to be 158 Mtonne CO2 (1988, excluding international bunkers and import) [9]. Of this amount 13 Mtonne CO2 was emitted by refineries, and 37 Mtonne CO2 by industry (excluding feedstock use of fuel). To obtain insight in the plants with large CO2 emissions, one should look at the lowest level of aggregation, i.e. per plant. Estimates of the combined combustion and process CO2 emissions resulted in a list of the top 20 Dutch CO2-producing industrial plants. These 20 plants are grouped per industrial sector in Figure 1. From Figure 1 it is clear that the largest CO2 emitting plants in the Netherlands are to be found among the refineries, in the basic metal industry, and in the petrochemical and fertilizer industries. It is worth noting here that the heavy industries are 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 already been described by Bakker [11]. In refineries gas streams with a high CO2 concentration can be identified in hydrogen manufacturing units and in residue gasification plants. Bakker calculated that for a residue gasification plant with a CO2 production of 2100 tonne per day, the CO2 can be recovered at a cost of U.S.$13 per tonne of CO2 avoided

442

JACCO C. M. FARLA ETAL. Mt C O 2 10~ 9~

"

87~ 65432 ~ 1 0

refineries

iron/steel

petrochem,

fertilizer

other chem.

chlorine

cement

Fig. 1. CO2 productionby the main industrial plants in the Netherlands in 1986. Each rectangle represents one industrial site [10]. [11]. However, this is only possible when these (planned) units are introduced in the Dutch refineries. In this study, we deal with carbon dioxide removal in the fertilizer industry, the iron and steel industry and the petrochemical industry.

3. The Fertilizer Industry The fertilizer industry is an important energy consumer worldwide. An important step in the manufacturing of nitrogen-fertilizers is the production of ammonia. In the Netherlands, the production of ammonia is responsible for over 90% of the energy consumption in the fertilizer industry. Ammonia is produced from carbon-containing feedstocks. The feedstocks worldwide are mainly natural gas, naphtha, coke or coal [12]. Ammonia does not contain any carbon, and carbon removal is an important step in all of the different production processes. Based on the worldwide ammonia production, and considering the different feedstocks, we estimate that over 220 Mtonne of carbon dioxide is recovered annually from ammonia production processes. Often, the carbon is recovered in the form of nearly pure carbon dioxide. This carbon dioxide may directly be compressed and transported to a storage site to prevent it from entering the atmosphere. Part of the recovered CO2 is used, the remaining is vented to the atmosphere. An important application of CO2 is the production of urea. The recovered CO2 is also used in the food and beverages industries and for other purposes.

CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES

443

In the next subsections we evaluate the possibilities for CO2 recovery based on the Dutch fertilizer industries. 3.1.

DESCRIPTION OF THE PRODUCTION PROCESS

The production of nitrogen-fertilizers is a large-scale industrial activity in the Netherlands. The ammonia production in the Netherlands amounted to 3.65 Mtonne in 1988 [13]. Over 2.5% of the world's total ammonia production capacity is based in the Netherlands [14], at five production locations. The fuel and feedstock input for ammonia production consists entirely of natural gas. For the production of ammonia, natural gas is heated, desulphurized and reformed with steam, which yields CO and H2. After this first reforming step, air is mixed with the gas stream in a second section. In the second reactor the unconverted methane is partially oxidized to yield CO, H2 and H20. The air is also used as the nitrogen source for the ammonia production. The process gas stream, containing H2, CO, CO2, H20, N2 and traces of the raw material, is cooled down and shifted. In the shift conversion, the carbon monoxide is converted with steam into carbon dioxide. After the shift conversion, the CO2 is removed from the process stream by means of a chemical absorption process [15, 16]. After CO2 removal, the process stream is cleaned of remaining carbon compounds. The process stream is then compressed to a high pressure and introduced into the ammonia reaction loop. In this reactor the nitrogen and hydrogen react over catalysts to yield ammonia. 3.2.

CARBON DIOXIDE REMOVAL FROM AMMONIA SYNTHESIS

The amount of natural gas used per tonne of ammonia produced is 33.5 GJ (1988) [13]. Some 38% of the natural gas is used for external heating purposes, which gives rise to combustion CO2 in the flue gases. The remaining 62% of the natural gas is used as feedstock and for internal fuelling [17]. The carbon in this share of the natural gas consumption is already being recovered, and needs only to be compressed and liquefied for transport. The CO2 in the flue gases may also be recovered by using an absorption process. However, we will concentrate on the carbon dioxide that is already being recovered. Approximately 1.2 tonne of carbon dioxide is recovered during the production of 1 tonne of ammonia [18]. With an ammonia production of 3.65 Mtonne (1988), the amount of recovered CO2 is 4.2 Mtonne-CO2 per year. Part of the recovered CO2 is used, the rest is vented to the atmosphere. A part is used in the synthesis of urea. Approximately 750 kg of CO2 is used for the production of 1 tonne of urea [12]. Since the annual production of urea is estimated to be 830 ktonne [19], the corresponding use of CO2 amounts to 0.6 Mtonne-CO2. At several ammonia production locations part of the recovered CO2 is liquefied

444

JACCO C.M. FARLAET AL.

in order to be sold to C O 2 c o n s u m e r s (mainly the food and beverages industry). These CO2 liquefying plants have an estimated annual production of 0.6 MtonneCO2 [20]. Based on these estimates of recovery and use, the amount of CO2 that is recovered and subsequently vented to the atmosphere will be approximately 3 Mtonne-CO2 per year. 3.3. COMPRESSION AND LIQUEFACTION OF C O 2 After cooling, the CO2 recovered from ammonia synthesis is delivered at a pressure of 1.3 bar and a temperature of 35 °C [16]. The carbon dioxide is saturated with water (5% v/v) and contains traces of nitrogen (300 ppm), hydrogen (1500 ppm) [16] and traces of methane [20]. The recovered CO2 must be compressed to 110 bar, and dried to a water content of lower than 10 ppm. Clean-up for the permanent gas contaminants is not necessary because of their low concentration. The amount of CO2 to be liquefied varies per ammonia production site. Based on the ammonia production capacities at the various production sites in the Netherlands, it is calculated that the compression and liquefaction plants must have capacities of between 400 and 1000 ktonne-CO2/year. The investment and energy costs are calculated for a compression facility of 700 ktonne of CO2 per year. The carbon dioxide is compressed in a four-stage isentropic compression process. The compression ratio will then be 3.0 in each stage. Compressor manufacturer Sulzer [21] has calculated the energy costs of compressing carbon dioxide to 110 bar, based on off-the-shelf equipment. For the calculations of the compression energy it was assumed that the CO2 was cooled to 35 °C between compression stages. When energy losses in the gear system are included, the compression energy amounts to 393 kJ/kg-CO2. Most of the water will be removed during the first compression stages. After the second stage the water content of the CO2 will be very low (approx. 1000 ppm; based on data from: [22]). However, additional drying is necessary to meet the specifications. Therefore the CO2 passes through a drying tower containing a solid desiccant, where the CO2 is dried to a water content of 10 ppm. Per tonne of CO2, 0.8 kg of water is removed during dehydration. The desiccant is regenerated by passing hot CO2 through the drying tower. A total dehydration energy of 8 kJ/kgCO2 is assumed. After the last compression stage, the carbon dioxide is cooled to 20 °C with cooling water. For cooling to 10 °C a small refrigeration step is needed. The energy demand for the refrigeration step is estimated to be 8 kJ/kg-CO2, based on a coefficient of performance of 5. 3.4. ECONOMIC EVALUATION A preliminary cost estimate was made for the compression of CO2 recovered in the ammonia production process. The investment costs for the compressor are derived from data provided by a manufacturer [21]. An installation factor of 2 was applied

CARBONDIOXIDERECOVERYFROMINDUSTRIALPROCESSES

445

TABLEII Cost calculationfor the removal of CO2 relatedto ammoniaproduction Capacity

(ktonne-CO2/yr)

700

Energy consumption Electricityconsumption Steam consumption

(MJ/tonne-CO2) (MJ/tonne-CO2)

401 8

Investment

Total investment Annual costs Capital costs O&M costs Electricity Steam Total annualcosts Specific costs CO2 avoided Specific mitigationcosts

(million U.S.$) (million U.S.$) (million U.S.$) (million U.S.$) (million U.S.$) (million U.S.$) (ktonne-CO2/yr) (U.S.$/tonne-CO2 avoided)

10 0.7 0.3 3.9 0.0 5.0 650 8

to the investment costs. For the drying equipment the investment costs were derived from [23], converted with a scaling power factor of 0.7. The investment costs for the refrigeration step are so small that they are neglected in this evaluation. The cooling water supply to the intercooling sections 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. An overview of the cost calculations, for a plant with a capacity of 700 ktonne-CO2 per year, is presented in Table II. For a plant size of 700 ktonne-CO2/yr, the compression and dehydration costs are U.S.$ 8 per tonne of CO2 avoided. Extrapolation to plant sizes between 400 and 1000 ktonne of CO2 per year changes these costs only slightly. The specific mitigation costs may decrease with a small amount when the carbon dioxide is compressed to the minimum required pressure of 80 bar [24], instead of 110 bar. A further decrease of the costs is possible by lowering the temperature between the compression stages.

3.5. CONCLUSIONSAND DISCUSSION Approximately 4.2 Mtonne of CO2 per year is recovered as part of ammonia production in the Netherlands. Some 3 Mtonne of the recovered CO2 is estimated to be vented to the atmosphere. This CO2 can easily be compressed and transported for storage outside the atmosphere (although storage options are not considered in this study). Preliminary cost calculations show that the costs of compressing the

446

JACCO C. M. FARLA ET AL.

COa will be approximately U.S.$ 8 per tonne of CO2 emission avoided. This cost estimate is considerably lower than the cost estimate for the cheapest option for the recovery of CO2 from power plants: 17 U.S.$/tonne-CO2 emission avoided for a newly built integrated-gasifier combined-cycle plant (IGCC) with shift conversion [3].

4. The Iron and Steel Industry Steel production is a highly energy intensive process. World steel production may roughly be divided in two distinct production methods: the scrap-based electric arc furnace (EAF) method, and the integrated route. Integrated steel production involves reduction of iron-ore to pig iron in a blast furnace, and then conversion to crude steel, usually in a basic oxygen furnace (BOF). Conversion of pig iron is also possible in an open hearth furnace (OHF). Based on the worldwide production figures (1990) in iron and steel plants [25], we estimate that the equivalent of over 1.1 Gtonne-CO2 of carbon emissions is generated in the integrated route processes. A large part of this carbon is available in the gases produced at several production steps. The carbon may be recovered from these gas streams. To date, these gases are used as fuel in and outside the iron and steel plants. This application need not be hindered by carbon dioxide recovery. In the next subsections, we evaluate the possibilities of carbon recovery in the iron and steel industry in the Netherlands. 4.1.

DESCRIPTION OF THE DUTCH IRON AND STEEL PRODUCTION PROCESS

In the Netherlands, more than 90% of the steel production consists of primary steel from the BOF integrated process. The remaining 10% is produced in a scrap-fed EAF process. The primary steel is produced at one iron and steel works only: the Hoogovens Group at IJmuiden. In 1986 the Hoogovens Group produced approximately 5.3 Mtonne of crude steel. The Hoogovens Group is the largest single energy consumer in the Netherlands, its energy use being approximately 10% of the total Dutch industrial energy consumption. The associated production of carbon-containing gases is estimated to amount to 8 Mtonne-CO2 (1986) [10]. A large part of the COa is not emitted at the Hoogovens production site, because the carbon containing gas is sold for power production in a regional power station. In 1986, gases with a calculated carbon content of more than 3.6 Mtonne-CO2 were sold to this power station. 4.1.1. Coke Production

To produce pig iron, large amounts of coke are needed. Coke is produced in coke ovens by the carbonization of coking coal. By-products of the coking process are coke-oven gas (CO gas), breeze (coke dust), tar and benzole. The by-products are

CARBON DIOXIDERECOVERYFROM INDUSTRIALPROCESSES

447

TABLE III Typical values for the composition and heat of combustion of some generated gaseous fuels (dry basis) during primary steel production [28, 29] CO gas

BF gas

BOF gasa

Flue gas BOF

3 56 21 20 0.045

14 70 16 -

2 52 2 2 42 -

Hydrogen Nitrogen Oxygen Carbon monoxide Carbon dioxide Methane C2-C6 hydrocarbons Hydrogen sulphide

% (v/v) % (v/v) % (v/v) % (v/v) % (v/v) % (v/v) % (v/v) g/m3

59 3 5 1 28 4 8b

Heat of combustion (LHV)

MJ/m3

19.8

3

8.8

0.5

80% of the steel production capacity in the Netherlands is equipped with a BOF gas recovery system. b Before washing; washing efficiency may vary.

separated and cleaned in a b y - p r o d u c t processing plant. The coke is introduced into the blast furnace, where it functions both as a fuel and as a reducing agent. The c o m p o s i t i o n of the c o k e - o v e n gas is shown in Table III.

4.1.2. Blast Furnace In the blast furnace the iron oxides in iron-ore are reduced chemically, at high temperature, to yield pig iron. During operation the blast furnace is fed with coke, iron-ore and lime. The heat in the blast furnace is generated by partial combustion of the coke, by introduction of preheated air (wind) and by the combustion o f injected pulverized coal. In the blast furnace, huge amounts of blast furnace gas (BF gas) are generated, consisting mainly of CO, CO2 and N2. The BF gas production is estimated to be nearly 1500 m 3 [27] per tonne of crude steel. Electrical energy is recovered f r o m the pressure o f the blast furnace gas (2-3 bar) by means of expansion turbines. The c o m p o s i t i o n of the blast furnace gas is shown in Table III. The pig iron is drawn off at the b o t t o m o f the blast furnace and transported to the steel plant.

4.1.3. Steel Production In a b a s i c - o x y g e n furnace (BOF) the pig iron is transformed into crude steel by decreasing the carbon content f r o m approximately 4% to below 1%. This is done by blowing o x y g e n into the molten iron in a converter. In the converter the carbon is oxidized, mainly to carbon monoxide, and is r e m o v e d via a flue gas treatment system. In the flue gas treatment system the carbon m o n o x i d e can either be burned for steam raising or the carbon m o n o x i d e can be recovered and used as a fuel.

448

JACCO C. M. FARLA ET AL.

If the BOF gas is recovered, a movable skirt is attached to the converter. During steel production this skirt is lowered to prevent air from entering the flue gas system and burning the carbon monoxide. Since the production of steel is a batch process, BOF gas is produced at time intervals. The collected gas is stored in a gas container before it is used or sold. The composition of both the recovered BOF gas and the flue gas from the basic-oxygen furnace is shown in Table III. 4.2.

CARBON BALANCE OF THE PRODUCTION PROCESS

The input of carbon in the iron and steel production process can be attributed to the coking coal, lime and other fuels like natural gas and injection coal. During the production process three secondary gas flows evolve; coke-oven gas, blast furnace gas and basic oxygen furnace gas. Some 85% of the carbon introduced into the process is present at one time in these three gas flows; approximately 70% in the BF gas, 9% in the CO gas and 7% in the BOF gas. The balance is incorporated in the steel, in the slag, in the by-products (breeze, tar, etc.) and in emissions in other process stages. Recovery of the carbon from these gases is possible before or after combustion. It seems advantageous to recover the carbon before burning, because nitrogen in the combustion air lowers the carbon concentration in the flue gases. Furthermore, the gases are produced at a few sources only, whereas they are used in several processes. A possible disadvantage of recovery before combustion is the fact that each of the gases contains carbon in a variety of compounds. The composition of the gases is shown in Table Ill. Since approximately 70% of the total carbon input emerges in the blast furnace gas, either a CO or CO2, recovery of the carbon from this gas will be described. 4.3.

CARBON DIOXIDE RECOVERY FROM BLAST FURNACE GAS

At present, pig iron is produced at Hoogovens in two blast furnaces with capacities of 6000 and 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 pressure recovery in expansion turbines [30]. In these turbines the pressure of the BF gas is reduced to 1.12 bar, and the gas is distributed for use as a fuel. Several separation technologies are available for the recovery of carbon dioxide from gases. These include pressureswing absorption, membrane purification and chemical or physical absorption processes [31]. Because of the low CO2 partial pressure in the blast furnace gas, a chemical absorption method is required. Several absorbing agents can be chosen for chemical absorption, like amines and alkali carbonates. Activated MDEA (methyldiethanolamine) is a suitable solvent for the recovery of CO2 from blast furnace gas [32]. It is chosen because of its relatively low energy requirement. MDEA is a tertiary amine and is selective for H2S in the presence of CO2.

CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES

449

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 CO2 recovery, unless the remaining dust causes the absorption solution to foam. The H2S that is recovered together with the CO2 will give approximately 0.5% H2S in the product gas. If desired this small amount of H2S can be separated easily. For the recovery of carbon dioxide from BF gas the most economic plant design is a one-stage chemical absorption configuration [32], Because of the low CO2 partial pressure practically no CO2 can be flashed. The heat consumption is estimated to be 140-160 kJ/mole CO2 [32]. The electric energy consumption is estimated to be 29 kJ/kg of CO2 (derived from: [33]). The recovered carbon dioxide has to be compressed to the transportation pressure of 110 bar and dried to a water content of below 10 ppm. Compression with a four-stage compressor step has been described in more detail in section 3.3. For the compression of the CO2 the electrical power requirement depends on the inletpressure of the recovered CO2. In the absorption process, a pressure drop of 0.1 bar is assumed. The carbon dioxide is cooled to 30 °C in the intercooling sections of the compressors. The compression energy will amount to 266 kJ/kg CO2 (based on data from: [21]). Drying of the carbon dioxide is also described in more detail in section 3.3. The total dehydration energy is estimated to be 8 kJ/kg of CO2. Cooling of the carbon dioxide to 10 °C (transportation condition) will consume 8 kJ of electricity per kg of CO2. Due to the recovery of CO2 from the BF gas, and due to the pressure drop in the absorption tower, less power will be generated in the expansion turbines. If 90% of the carbon dioxide is recovered, the electrical power generation in the expansion turbines will fall from approximately 19 MW to 15 MW, which represents a loss of electricity production equivalent to 43 kJ per kg of CO2 recovered (on average for the two blast furnaces). 4.3.1. Shift Conversion The carbon recovery from blast furnace gas may be enhanced when the carbon monoxide in the gas is converted into carbon dioxide. This can be accomplished with the water-gas shift-reaction. Sufficient conversion of the carbon monoxide can be obtained in two serial reactors: a high-temperature shift reactor operating at approximately 400 °C followed by a low-temperature shift reactor operating at a temperature of over 250 °C. The shift reaction is followed by a CO2 recovery process. After conversion approximately 5.7 Mtonne-CO2 can be recovered. This option will not be further elaborated, but it is estimated that this would result in a higher price of recovery, because of the heat required for the shift reaction and because of the extra investment costs.

450

JACCO C. M. FARLAET AL. TABLE IV Cost calculations for the recovery of CO2 from blast furnace gas Capacity Energy consumption Electricity consumption Loss of electricity production Steam consumption Investment Total investment Annual costs Capital costs O&M costs Electricity Steam Loss of electricity production Total annual costs Specific costs CO2 avoided Specific mitigation costs

(ktonne-CO2/yr)

2800

(MJ/tonne-C02) MJ/tonne-COz) MJ/tonne-C02)

310 43 3417

million U.S.$)

373

million million million million million million

U.S.$) U.S.$) U.S.$) U.S.$) U.S.$) U.S.$)

(ktonne-CO2/yr) (U.S.$-tonne-C02 avoided)

27 8 12 24 2 72 2042 35

4.4. ECONOMIC EVALUATION

A preliminary cost estimate has been made for the recovery of CO2 from blast furnace gas. For the MDEA chemical absorption process, investment cost figures for an MEA process are used, on the basis of plant similarity. The economic evaluation is based on 8350 operating hours per year. Ninety per cent of the CO2 in the blast furnace gas will be recovered, resulting in 8 ktonne-CO2 per day. Since the maximum train capacity is of the order of 1000 tonne CO2 per day, 8 trains will be needed. The total capital costs for 8 ktonne-CO2 are estimated to be U.S.$ 352 million (based on data from [34]). Operation and maintenance costs per year are calculated to be 2.1% and 3.6% of investment costs, for static and rotating machinery respectively. The carbon dioxide should be compressed to 110 bar. This is accomplished with four-stage centrifugal compressors. The investment for the compressor is estimated to be U.S.$18 million, based on data from Smit [21]. The solid desiccant dehydration units are estimated to cost U.S.$ 3 million, based on data from KTI [23]. The investment costs for the refrigeration step are so small that they are neglected in this evaluation. The cost figures, for the two blast furnaces together, are presented in Table IV.

CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES

451

Carbon dioxide may be recovered from blast furnace gas and compressed to 110 bar at a cost of approximately U.S.$ 25 per tonne of carbon dioxide. When the extra CO2 emissions due to steam and electricity use are accounted for, one arrives at a cost figure of U.S.$ 35 per tonne of CO2 emission avoided. If low-temperature waste heat can be used for the recovery of CO2, the costs per tonne of CO2 avoided will become lower. We estimate that the costs can fall by one quarter if waste heat can be used (see also section 5.3). However, no large waste heat streams have been located in the iron and steel production processes. 4.5. CONCLUSIONS AND DISCUSSION

Carbon dioxide can be recovered from several secondary fuel gases in the iron and steel industry. The largest part of the carbon introduced into the iron production process evolves in the blast furnace gas. Therefore, the possibility of recovering CO2 from blast furnace gas has been assessed. A preliminary cost estimate for the recovery of CO2 from blast furnace gas is U.S.$ 35 per tonne of CO2 emission avoided. This cost estimate indicates that the recovery of CO2 from blast furnace gas is expensive, and comparable to the cost of recovering CO2 from stack gases of coal-fired thermal power plants with a chemical absorption process, which is also U.S.$ 35 per tonne of CO2 avoided [3].

5. The Petrochemical Industry The petrochemical industry consumes large amounts of energy worldwide. In the petrochemical industry natural gas and oil-derived feedstocks are converted to organic bulk chemicals. These bulk chemicals, like e.g. ethylene, propylene, butadiene and aromatics are processed further within or outside the petrochemical industry (e.g. in the resins and synthetics industry). The largest energy consuming unit process is the (steam) cracking of the initial feedstock. Carbon dioxide may be recovered from flue gases in the petrochemical industry. There are no worldwide CO2 emission figures, but we estimate that petrochemical emissions are of the order of 1 Gtonne CO2 per year. The petrochemical industry is the largest energy consumer of the Dutch chemical industry sector, with a fuel consumption of over 300 PJ in 1985 (including feedstocks) [35]. The main fossil input in the petrochemical industry consists of naphtha and gasoil. The CO2-emission of the Dutch petrochemical industry (combustion emissions) was 8.5 Mtonne CO2 in 1986 [36]. The three largest petrochemical industries in the Netherlands depicted in Figure 1, are estimated to emit 7.3 Mtonne of CO2 annually (including process emissions). In the next subsections, the possibilities of CO2 recovery in the Dutch petrochemical industries are assessed.

452 5.1.

JACCO C. M. FARLA ET AL. SHORT DESCRIPTION OF THE PRODUCTION PROCESSES

The feedstock for the Dutch petrochemical industry is derived from refineries. The hydrocarbon feed is cracked at high temperatures (850 °C). The cracked gaseous products, ranging from hydrogen to heavy fuel oil, are quenched. During quenching an aromatic fraction (pyrolysis gasoline) is separated which is used as a raw materia for the production of aromatics. The cracked gas is compressed and fractionated tc yield, among other products, ethylene, propylene and butadiene. The production of ethylene, propylene, butadiene and benzene accounted for over 80% of the total fuel consumption in the petrochemical industry in 1986 [35]. The CO2 emissions associated with these products are combustion emissions. In sub-section 5.2 we will assess the CO2 recovery possibilities from these emissions. The production of ethylene oxide (EO) accounted for only a very small fraction of the total fuel consumption in the petrochemical industry in 1986 [35]. However, the production olEO is described briefly because a highly concentrated CO2-stream is separated during production. Ethylene oxide is produced by direct oxidation of ethylene with oxygen over a silver catalyst. Per tonne of ethylene oxide approximately 0.88 tonne of CO2 is formed as a by-product [12]. In 1981 the production capacity of ethylene oxide in the Netherlands was 325 ktonne [37]. The amount of recovered CO2 associated with ethylene oxide production is nowadays estimated to be of the order of 250 ktonne/yr. This CO2 contains only traces of methane, ethylene and ethylene oxide [38]. After recovery, the only steps needed will be dehydration and compression. 5.2.

CARBON DIOXIDE RECOVERY IN THE PETROCHEMICAL INDUSTRY

As indicated in subsection 5.1, the C O 2 emissions associated with the steamcracking of naphtha are combustion emissions. The furnace and steam boilers are fueled with the by-products of the fractionation steps. To be able to estimate the CO2 content of the flue gases, one needs to know the composition of the fuel mix. Chauvel and Lefebvre [12] show that the gaseous fuel production is about 18% (m/m) of the naphta input. Some 5% (m/m) of the naphta input is converted into heavy products which are also used as fuel. If the gaseous fuels are burned in a gas turbine (e.g. in combined heat and power production), the CO2 concentration in the flue gases will be low (~ 4%), because a gas turbine is fired with excess air. With such low CO2 concentrations, CO2 recovery will be relatively expensive. If the gaseous fuels and the heavy fuel products are burned atmospherically for furnace heating, it is estimated that the CO2 concentration ranges from 7-10% (v/v) and carbon dioxide recovery from the flue gases will be more feasible. On the basis of data in [39] it is estimated that at present less than 10% of the fuel in the petrochemical industry is burned in (CHP) gas turbines. Therefore it may be possible to recover the CO2 from 90% of the stack gases at the three big petrochemical sites. This corresponds to 6 Mtonne of CO2 per year.

CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES

453

The C O 2 from the stack gases can be recovered by means of a chemical absorption process. This process will be described for a plant with a capacity of 2 Mtonne CO2 per year, which corresponds to the CO2 production at each of the three large petrochemical industries. Because of the low CO2 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-CO2 [34]. The electricity consumption of the absorption process is estimated to be 29 kJ/kg-CO2. The flue gases are compressed with blowers to compensate for the pressure drop in the absorption column. The pressure drop is assumed to be of the order of 0.1 bar, and the electric energy consumption for the stack gas blowers is then calculated to be approximately 72 kJ/kg of CO2 (based on a CO2 concentration of 10% in the flue gases). After recovery, the carbon dioxide has to be compressed for transport. Compression is described in some detail in section 3.3. The compression energy required for compression to 110 bar will be 370 kJ/kg of CO2 (derived fi'om data by Smit [21]). The carbon dioxide is dried by passing it through a solid desiccant drying tower. The total dehydration energy is estimated to be 8 kJ/kg of CO2. Cooling of the carbon dioxide to a temperature of 10 °C will consume 8 kJ of electricity per kg CO2. 5.3.

ECONOMIC EVALUATION

A preliminary cost estimate has been made for the recovery of C O 2 from the stack gases in the petrochemical industry. The evaluation is based on the CO2 recovery units being on-stream for 8000 hours per year. 90% of the CO2 in the stack gases will be recovered, which means approximately 5.4 ktonne CO2 per day for each of the three large petrochemical plants. Because the maximum train capacity is of the order of 1000 tpd, six trains will be needed. The total installed capital costs for 5.4 ktonne/day recovery are estimated to be U.S.$ 238 million (based on data from [34]). The cost estimate includes stack gas cooling equipment and stack gas blowers. Operation and maintenance costs per year are calculated to be 2.1% and 3.6% of investment costs, for static and rotating machinery respectively. The carbon dioxide should be compressed to 110 bar. This is accomplished with a four-stage compressor. The investment for the compressor is estimated to be U.S.$12 million, based on data from Smit [21]. The solid desiccant dehydration units are estimated to cost U.S.$ 2.4 million, derived from data by KTI [23]. The investment costs for the refrigeration step are so small that they are neglected in this evaluation. The cost figures are presented in Table V. In the preliminary cost estimate a price of U.S.$ 46 per tonne of CO2 avoided is calculated. This price is based on a steam price of 2.5 U.S.$/GJ and a steam-related CO2 emission factor of 62 kg-CO2/GJ. A sensitivity analysis has been performed

454

JACCO C.M. FARLAET AL. TABLE V Cost calculation for the recovery of CO2 from flue gases in the petrochemical industry Capacity

(ktonne-CO2/yr)

1800

Energy consumption Electricity consumption Steam consumption

(MJ/tonne-CO2) (MJ/tonne-CO2)

479 4208

Investment

Total investment Annual costs Capital costs O&M costs Electricity Steam Total annual costs Specific costs CO2 avoided Specific mitigation costs

(million U.S.$)

252

(million U.S.$) (million U.S.$) (million U.S.$) (million U.S.$) (million U.S.$)

18 6 12 19 54

(ktonne-CO2/yr) (U.S.$/tonne-CO2 avoided)

1178 46

TABLE VI Specific mitigation costs for CO2 from stack gas (in U.S.$/tonne of CO2 avoided) with varying steam prices and CO2 emission factors Steam price (U.S.$/GJ) CO2 emission factor of steam (kg/GJ)

3.5 2.5 1.5

62

31

53 46 40

39 33

by varying the steam-price and the CO2 emission factor assigned to steam. The results of this sensitivity analysis are given in Table VI. Comparable to the situation in the iron and steel industry, lower costs per tonne o f CO2 avoided can be obtained when waste heat is used. During this study no waste heat streams were identified in the petrochemical industry. The total heat consumption of the three large petrochemical plants is calculated with data in [39]. The energy consumption for recovery of the CO2 from the stack gases amounts to 2 0 - 3 0 % of the total heat consumption, which is equivalent to approximately 1 0 - 1 5 % o f the total fuel and feedstock demand.

CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES

455

Shell Chemicals [38] reported that the cost of LP steam (3 bar, saturated) is 1.6 U.S.$/GJ when fired with oil, and 2 U.S.$/GJ when fired with natural gas, based on current fuel prices. On the basis of these prices, the cost of recovery will be lower than U.S.$ 45 per tonne of CO2 avoided, and may arrive - depending on the attributed CO2-emission - at a typical price of U.S.$ 38 per tonne of CO2 avoided. The carbon dioxide which is co-produced in the manufacture of ethylene oxide need only be compressed and dehydrated. Ethylene oxide is produced at 3 sites in the Netherlands. Per location a CO2 liquefaction unit with a capacity of the order of 80 ktonne-CO2/yr may be built. The cost per tonne of CO2 avoided will be about U.S.$9. 5.4.

CONCLUSION AND DISCUSSION

Carbon dioxide can be recovered from the stack gases of petrochemical processes. A preliminary cost estimate for the recovery of CO2 from these stack gases yields a price of approximately 45 U.S.$/tonne of CO2 avoided. This cost estimate indicates that recovery of CO2 in the petrochemical industry is expensive, even more costly than the cost of recovering CO2 from stack gases of a coal-fired power plant equipped with a chemical absorption process, which is approximately 35 U.S.$/tonne CO2 [3]. If low-temperature waste heat can be recovered and used for the recovery of CO2, the costs per tonne of CO2 avoided may fall to some extent. The cost of compressing the carbon dioxide which has already been recovered during the manufacture of ethylene oxide is considerably lower, and is estimated to be 9 U.S.$/tonne of CO2 avoided.

6. Importance for Other Countries In the previous sections we have evaluated the possibilities for recovering part of the carbon dioxide emissions from three branches of industry in the Netherlands. If the CO2 is recovered, as described in the previous sections, approximately 5% of the national carbon dioxide emissions can be avoided. How is the potential for CO2 recovery in these sectors for other countries? In order to give a preliminary answer to this question, we will give an estimate of the potential for CO2 recovery in the European Union (E.U.), which can be considered as being more or less representing mature industrialized societies. In Table VII we give an overview of the estimated CO2 emissions in the three industrial sectors for the E.U. The total CO2 emissions in the European Union are approximately 2.7 GtCO2/yr (derived from [41, 42]). Thus, the CO2 emissions given in Table VII are equal to 13% of the CO2 emission in the European Union.

456

JACCO C.M. FARLA ET AL.

TABLE VII Overviewof the carbondioxideemissionin three industrial sectors in the European Union (1988) [40-45] Industrial sector

C02 emission

(Mtonne C02) Ammonia production (feedstock) 45 Iron and steel industry 220 Petrochemical industry (fuel) 90

As in the Netherlands, most of the ammonia production in the E.U. is using natural-gas based methane reforming, followed by CO2 removal and ammonia synthesis. Only about 8% of the ammonia production is based on partial oxidation of hydrocarbons [ 13]. For the possibilities of CO2 recovery, both production methods do not differ to a large extent. The iron and steel production processes are more or tess comparable to those in the Netherlands. The share of steel making by electric arc furnaces (out of secondary steel) is larger in other countries than in the Netherlands; however this process mainly consumes electricity. In all countries the CO2 emission can mainly be allocated to the blast furnace/basic oxygen furnace route [25]. No data were available on the gas composition of the blast furnace gases in other countries, but there is no reason to expect that the volumes and compositions will deviate strongly. In the petrochemical industry we have assumed that CO2 will be recovered from flue gases. This means that the exact processes applied will not be very important for the CO2 recovery assessment. Altogether we may well assume that the fractions of CO2 to be recovered may be the same as for the Netherlands. By applying the recovery processes described in this paper, approximately 4% of the CO2 emissions in the E.U. may be avoided.

7. D i s c u s s i o n

It should be emphasized that this study is meant as an exploratory investigation into the possibilities of removing CO2 from large-scale industrial processes. The amount of recoverable CO2 and the unit recovery costs depend on the actual processes in an industrial plant. We took the situation in the Netherlands as an example because of data availability. One should be aware of the fact that layout and operational considerations (e.g. start up problems) were not part of this study. Also, no intermediate CO2 storage

CARBON DIOXIDE RECOVERY FROM INDUSTRIAL PROCESSES

457

capacity was taken into account, although transport per pipeline may require an uninterrupted supply of CO2 to the pipeline. Furthermore, it is important to realise that current or future changes in production processes and product mixes will influence the CO2 recovery possibilities described in this study. Generally, we may expect an increasing energy efficiency in industry. This development may decrease the amount of CO2 that can be recovered. A further increase of the industrial energy efficiency may be attained by enlarging the capacity of combined heat and power production (CHP). If gas turbines are used in this new CHP capacity, CO2 recovery will become much more expensive because of the low CO2 concentrations in gas turbine exhaust gases.

8. Conclusions

"

Large amounts of carbon dioxide are emitted by industry worldwide. This carbon dioxide may be recovered and then stored outside the atmosphere. In this study a preliminary assessment is made of the potential and the costs of CO2 recovery in industry, based on the industrial situation in the Netherlands. The industries with the highest CO2 emission in the Netherlands are the refineries, the fertilizer industry, the iron and steel industry and the petrochemical industry. During ammonia production, carbon dioxide is already r~ecovered as part of the process. For the ammonia production, it is estimated that per year 2.8 Mtonne of carbon dioxide emission can be avoided at a cost of 8 U.S.$/tonne of CO2. In the iron and steel industry, many energy conversion processes take place. Very large conversion processes take place during coke production and in the blast furnaces. In the blast furnaces large volumes of blast furnace gas are formed, containing 20% (v/v) CO2. This carbon dioxide can be recovered from the BF gas. In this way the emission of 2 Mtonne of CO2 can be avoided at an estimated price of U.S.$ 35 per tonne. The price could become lower if waste heat can be used in the recovery process. Carbon dioxide can also be recovered from the stack gases in the petrochemical industry. In this study, it is estimated that approximately 5.4 Mtonne of CO2 can be recovered annually. Because of the extra CO2-emissions relating to recovery and compression, this will result in 3.5 Mtonne of CO2 avoided, at a cost of approximately 45 U.S.$/tonne of CO2 avoided. This cost estimate may become~ lower t o o i f waste heat can be used. t n the petrochemical industry one small high concentration CO2 source has been identified, related to the ethylene oxide production. It is estimated that from this source, over 0.2 Mtonne of carbon dioxide emission can be avoided per year at a cost of 9 U.S.$/tonne of CO2. :," ][1
458

JACCO C. M. FARLA ET AL. 8"

coal-fired thermal power plants

natural gas-fired thermal powerplants

7" 6"

~'5-

Petrochemical industry

~4-

oo3.

Ammonia Iron and Steel oxide

2"

1

~

[

J

i

5

i

i

10

L

Refinades ~

i

i

i

J

i

i

I

i

i

15 20 25 30 35 40 Specific mitigation costs U.S.$/ton-CO2)

CO2 recovered ~

CO2 avoided [ ~

L

i

b

45

L

50

Thermal power plants

Fig. 2. Amount of recoverableand avoidableCO2 in relationto the specificmitigationcosts [46].

emission can be avoided if carbon dioxide is recovered from refinery flue gases, and if the blast furnace gas is shifted before CO2 recovery. The CO2 recovery possibilities in the European Union were assessed briefly for the same branches that were studied for the Netherlands. We found no reason to assume that the CO2 recovery possibilities in the E.U. are different from the recovery possibilities in the Netherlands. In extrapolating the results for the Netherlands to the E.U., we found that nearly 4.5% of the CO2 emissions can be avoided if CO2 recovery is applied in the three heavy industry branches that we studied. The cost figures for recovery and compression of CO2, as reported in this study, can be compared with cost figures for CO2 recovery from power production plants. The lowest cost figure reported by Hendriks et al. [3] is U.S.$17 per tonne of CO2 avoided for a newly built IGCC with shift conversion of the fuel gas. This cost figure is considerably higher than the cost of recovery and compression of CO2 from the ammonia and ethylene oxide production. Recovery of CO2 from the flue gases of a thermal power plant is reported to cost U.S.$ 35 and U.S.$ 40 per tonne CO2 avoided, for coal-fired and natural gasfired power plants respectively [3]. These cost figures are comparable with the cost figures for the recovery of CO2 in the iron and steel industry, but lower than the estimated cost of recovery of CO2 from the stack gases in the petrochemical industry. In Figure 2 the amount of CO2 that can be recovered is depicted together with the cost of recovery, separately for the industrial sectors studied. More detailed studies will be necessary to identify problems not recognized in this study or to find possibilities for CO2-remdval at lower cost, including adapti0n

CARBON DIOXIDERECOVERYFROM INDUSTRIALPROCESSES

459

o f industrial processes in such a way that CO2 recovery can be p e r f o r m e d m o r e easily.

Acknowledgements This study forms part of the 'Integrated Research P r o g r a m m e on Carbon Dioxide R e m o v a l and Storage' (SOP-CO2). The SOP-CO2 p r o g r a m m e was supported financially by the Dutch Ministry o f Housing, Physical Planning and the Environment, Directorate-General for Environmental Protection ( V R O M - D G M / H D M / L E ) and by the P r o g r a m m e C o m m i t t e e of the Dutch National Research P r o g r a m m e on Global Air Pollution and Climate Change ( N O P - M L K ) . The authors would like to thank Ernst Worrell for his valuable c o m m e n t s on an earlier draft of this paper. The authors are also grateful to manufacturers and industries who supplied us with some indispensable data.

References 1. Houghton, J. T., Jenkins/G. J., and Ephraums, J. J. (eds.): 1990, Climate Change: The IPCC Scientific Assessment, Cambridge University Press, Cambridge, U.K. 2. Blok, K., Turkenburg, W. C., Hendriks, C. A., and Steinburg, M. (eds.): 1992, Proceedings of the First International Conference on Carbon Dioxide Removal, Amsterdam, Pergamon Press, Oxford, U.K. 3. Hendriks, C., Farla, J., and Blok, K.: 1992, Verwijdering en Opslag van C02 bij Elektriciteitsopwekking (Recovery and Storage of CO2from Electricity Production ), Report 92035, Utrecht University, Dept. of Science, Technology and Society, Utrecht, The Netherlands (in Dutch). 4. Costs were calculated in Dutch Guilders of 1990. For convenience, costs are reported in U.S. Dollars by dividing the calculated cost figures by 2 (exchange rate of begin 1990). 5. Heijningen, R. van (ed.): 1990, Reductie van de Kooldioxide-Uitstoot via het AfvaIstoffenbeleid (Reduction of Carbon Dioxide Emissions by Waste Management Policy), Report 140346.01, Ministry of Housing, Physical Planning and Environment VROM, The Hague, The Netherlands (in Dutch). 6. All efficiencies are expressed on a lower heating value (LHV) basis. 7. Houghton, J. T., Callander, B. A., and Vamey, S. K. (eds.): 1992, Climate Change 1992: The Supplementary Report to the IPCC Scientific Assessment, Cambridge University Press, Cambridge, U.K. 8. OECD/IEA: 1993, Energy Statistics and Balances of Non-OECD Countries 1990-1991, Report 61-93-13-3, Paris, France. 9. RIVM: 1991, Nationale Milieuverkenning 2 (National Environmental Outlook 2), National Institute of Public Health and Environmental Protection RIVM, Samsom H. D. Tjeenk Willink B.V., Alphen a/d Rijn, The Netherlands (in Dutch). 10. Data calculated using a database that was developed in our department (for details: see Ref. [26]). The database contains figures (base-year 1986) concerning the energy consumption of the 296 industrial plants with the largest energy consumption in the Netherlands. 11. Bakker, H.: 1992, C02 Emission from Refineries, Report 60906-01, Comprimo Engineers and Contractors, Amsterdam, The Netherlands. 12. Chauvel, A. and Lefebvre, G.: 1989, Petrochemical Processes, Institut Frangais du P6trole Publications, l~ditions Technip, Paris, France. 13. Worrell, E. and Blok, K.: 1994, 'Energy Savings in the Nitrogen Fertilizer Industry in the Netherlands', Energy 19 (2), 195-209.

460

JACCO C. M. FARLA ET AL.

Constant, K. M. and Sheldrick, W. E: 1991, An Outlook for Fertilizer Demand, Supply, and Trade, 1988/89-1993/94, Report 137 (World Bank Technical Paper), The World Bank, Washington, D.C., U.S.A. 15. Blanken, J. M.: 1988, 'De Ammoniakindustrie (The Ammonia Industry)', i2-Procestechnologie 2, 37-42 (in Dutch). 16. Versteele, W.: 1992, Written Communication, Hydro Agri Sluiskil, B.V., Sluiskil, The Netherlands. 17. Flint, J. G. Th., Heddema, E J., and Lokerse, R J.: 1990, Energiebesparingspotenti~Ten; Ontwikkelingen 1986-2015 voor de Chemische Industrie (Energy Conservation Potentials; Developments 1986-2015for the Chemical Industry), Erbeko, Hilversum, The Netherlands (in Dutch). 18. Calculated with the CO2 emission factor of 'Groningen' natural gas of 56.1 kg-CO2/GJ. 19. Worrell, E.: 1992, Personal Communication, Utrecht University, Dept. of Science, Technology and Society, Utrecht, The Netherlands. 20. Woudstra: 1992, Personal Communication, Kemira B.V., Corporate Research and Development, Vondelingenplaat (Rotterdam), The Netherlands. 21. S mit, J. B.: 1992, Written Communication, Sulzer Nederland B.V., Zoetermeer, The Netherlands. 22. Song, K. Y. and Kobayashi, R.: 1987, Water Content of CO2 in Equilibrium with Liquid Water and~or Hydrates, SPE Formation Evaluation, December 1987, 500-508. 23. KTI: 1992, Removal of C02 from Reformer Gas in a Power Plant, Report VROM 262.848, Kinetics Technology International, Zoetermeer, The Netherlands. 24. Hendriks, C. A. and Blok, K.: 1993, 'Underground Storage of Carbon Dioxide', Energy Convers. Managem. 34, (9-11), 949-957. 25. IISI: 1992, Steel Statistical Yearbook 1992, International Iron and Steel Institute, Brussels, Belgium. 26. Blok, K. and Worrell, E.: 1992, 'Heat and Electricity Consumption of Large Industrial Energy Users in the Netherlands', Heat Recov. Syst. CHP 12, (5), 407417. 27. Volumetric figures are given in m 3 at standard pressure and temperature (1.013 bar, 0 ° C). 28. VROM: 1988,HandbookofEmission Factors, Report VROM 7102512-88,Ministry ofHousing, Physical Planning and Environment VROM, The Hague, The Netherlands. 29. 'Hoogovens Gaat Koolmonoxide van Staalfabriek Benutten (Hoogovens Will Use Carbon Monoxide from Steel Plant', Energie MilieutechnoI. 1991, 1, 10-11 (in Dutch). 30. Wolthuizen, S.: 1988, IJzel, Staal en Walstechnologie (Iron, Steel and Casting Technology), Hoogovens IJmuiden, IJmuiden, The Netherlands (in Dutch). 31. Oudhuis, A. B. J.: 1992, Inventarisation of Techniques for C02 RemovaI fivm Fuel Gas or Reformer Gas, Report ECN-C-92-043, Netherlands Energy Research Foundation ECN, Petten, The Netherlands. 32. Vannby, R.: 1992, Written Communication, ref. 20100RV/NM, Haldor TopsCe A/S, Lyngby, Denmark. 33. Gerhardt, W. and Hefner, W.: 1988, BASF's Activated MDEA - A Flexible Process to Meet . Speeiflc Blant Co~tditions, in AIChE Ammonia Safety Symposium, Denver, CO, U.S.A. 34. Mariz, C.: 1991, Written Communication, Fluor Daniel, Inc., Irvine, CA, U.S.A. 35. Worrell, E., de Beer, J. G., Cuelenaere, R. E A., and Blok, K.: 1992, ICARUS, The Potential for Energy Conservation in the Netherlands up to the Year 2000. Report 92024. Utrecht University, Dept. of Science, Technology and Society, Utrecht, The Netherlands. -~ ,36:: " va/]~e~Hbut "K, 'D ~~d.) i 1990 l~dustri~le Emissies in uederland; Derde Inventarisa~ieronde :. ,;' ~ !~5"~:19877;(I6~r~i3~L :Eniisgions ~in the:~Netl,terlands; :Th!rd lnvento~ 1985-1987), Public:-:, ....catie~ee~s Emi@ei~gistr~ttie, Mini-stry of Hous~ng~Physical Planning and Environment VROM, " The Hague, The*Ne,~herlands (iri.Dutch)., . ~ . ":3). Cl~mfact's: ~t9~1 ;~herhfactsNethe?landL Chemical Data Seryices. IPC Industrial Press, Sutton, Surrey~,:,U,:K.. ~.2 '" ,:. . . . . . . . . . i 38. Bergh~ 13. R. ).d.-'1992. Written Communication,ref. MAT (mat21.92), Shell Nederland Chemie b.v., Moerdijk,'The Netherlands.39:: ~: TNO: 1986,,Em " i sste . . Reg, . . tstratteLucht . . . . . . (Emtsston . . . . Regtstratton . . . . . . Atr . .PolIutton p. TNO Environ-:.~ r/a~at an~t,Eni~gy R~search, Delft(Apel~iookn,~The ,NettaerlaJads~(inDutch) 14.

"

CARBON DIOXIDERECOVERY FROM INDUSTRIALPROCESSES 40.

41. 42. 43.

44. 45.

461

Data on the 'iron and steel' and 'chemical' industry are taken from Ref. [41] (Annex 1) and Ref. [42] by adding the carbon emissions of the 12 countries of the European Union (the former European Community). Data on 'France' were derived from Ref. [42]. Carbon emission factors were taken from Ref. [41]. OECD/IEA: 1991, Energy Efficiency and the Environment, Energy and the Environment Series, OECD/IEA, Paris, France. OECD/IEA: t990, Energy Balances of OECD Countries 1987/1988, OECD/IEA, Paris, France. Figures on fuel consumption for ammonia production in the European Union were taken from Ref. [44]. We assume a feedstock/fuel ratio of 60/40 (derived from Ref. [ 13]). Carbon emission factors were taken from Ref. [41]. Worrell, E., Cuelenaere, R. E A., Blok, K., and Turkenburg, W. C.: 1994, 'Energy Consumption by Industrial Processes in the European Union', Energy 19, (11), 1113-1129. Data on the petrochemical industry alone are not available. We first subtracted the fuel and feedstock consumption for ammonia production (Ref. [43]) from the fuel and feedstock consumption of the 'chemical industry' (Ref. [42]). The remaining feedstock consumption of oil is completely attributed to the 'petrochemical industry'. Furthermore we assumed the fuel consumption of the 'petrochemical industry' to be equal to the feedstock consumption (Ref.

[35]). 46.

47.

Besides the data calculated in this paper, the data on recovery of carbon dioxide from thermal electric power plants (coal and natural gas fired) are added for reference. These cost figures are taken from Ref. [3]. The input of coal and natural gas in the production of electricity (1990) are taken from Ref. [47]. The amounts of CO2 emission from electricity production that can be avoided, are 18.7 and 11.5 Mtonne-CO2, for coal fired and natural gas fired plants respectively. CBS: 1991, De Nederlandse Energiehuishouding: Jaarcijfers (Energie-SuppIy in the Netherlands: Annual Figures), Vol. 1990, Netherlands Central Bureau of Statistics, Voorburg, The Netherlands (in Dutch).