Journal of Environmental Management 132 (2014) 1e8

Contents lists available at ScienceDirect

Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman

Deciding between carbon trading and carbon capture and sequestration: An optimisation-based case study for methanol synthesis from syngas
a, *, Semra Ag
ralı b, Yıldız Arıkan a, Eray Avcıog
lu c Fehmi Görkem Üçtug
an Caddesi, 34353 Besiktas, Istanbul, Turkey Bahçes¸ehir University, Department of Energy Systems Engineering, Çırag Bahçes¸ehir University, Department of Industrial Engineering, 34353 Besiktas, Istanbul, Turkey c Politecnico di Milano, Management Engineering, Via Anzani 9, Como, Italy a

b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 November 2012 Received in revised form 11 July 2013 Accepted 21 October 2013 Available online 16 November 2013

The economic and technical feasibility of carbon capture and sequestration (CCS) systems are gaining importance as CO2 emission reduction is becoming a more pressing issue for parties from production sectors. Public and private entities have to comply with national schemes imposing tighter limits on their emission allowances. Often these parties face two options as whether to invest in CCS or buy carbon credits for the excess emissions above their limits. CCS is an expensive system to invest in and to operate. Therefore, its feasibility depends on the carbon credit prices prevailing in the markets now and in the future. In this paper we consider the problem of installing a CCS unit in order to ensure that the amount of CO2 emissions is within its allowable limits. We formulate this problem as a non-linear optimisation problem where the objective is to maximise the net returns from pursuing an optimal mix of the two options described above. General Algebraic Modelling Systems (GAMS) software was used to solve the model. The results were found to be sensitive to carbon credit prices and the discount rate, which determines the choices with respect to the future and the present. The model was applied to a methanol synthesis plant as an example. However, the formulation can easily be extended to any production process if the CO2 emissions level per unit of physical production is known. The results showed that for CCS to be feasible, carbon credit prices must be above 15 Euros per ton. This value, naturally, depends on the plant-specific data, and the costs we have employed for CCS. The actual prices (z5 Euros/ton CO2) at present are far from encouraging the investors into CCS technology. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: CO2 emission reduction CCS Carbon trading GAMS Non-linear optimisation Methanol

1. Introduction In recent decades, increasing greenhouse gas (GHG) emission is one of the main reasons for global warming with adverse environmental effects such as sea level rise, floods, droughts, etc. Fossil fuel power plants, and many other industries such as iron and aluminium, cement, lime, hydrogen, ammonia and methanol plants have been known as the major emission sources of greenhouse gases, of which the most abundant one in the atmosphere is carbon dioxide, or simply CO2 (IPCC Fourth Assessment Report, 2007; Soltanieh et al., 2012). Industrial activities account for 40% of global energy-related CO2 emissions. In 2007 the worldwide figure for CO2 emissions caused by industrial activities was 7.6 gigatonnes equivalent (Gte) of direct CO2 emissions whereas the indirect CO2 emissions due to electricity production for industrial activities were * Corresponding author. Tel.: þ90 (0) 212 381 5691; fax: þ90 (0) 212 381 0550.
). E-mail address: [email protected] (F.G. Üçtug 0301-4797/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jenvman.2013.10.016

3.9 Gte (Roddy, 2012). Reduction of fossil fuels (oil, gas, coal) consumption via enhanced energy efficiency, carbon capture and sequestration (CCS), conversion of CO2 to different products, increasing utilisation of renewable energies and reforestation are effective options to achieve mitigation so as to reach the Kyoto Protocol targets (Chicco and Stephenson, 2012). Coal-fired power plants are the most common source of electricity production with a global share of 41%. Despite the high level of environmental impact of coal combustion, economic incentives urge many countries towards a coal-dominant energy sector, examples of such countries being U.S.A., China, India and partly Turkey (Cristóbal et al., 2012b). Turkey produces approximately 25% of its electricity through local or imported coal (Capik et al., 2012), and coal is the main energy source of the country as far as local resources are concerned, a situation caused by inadequate natural gas and petroleum reserves. Hence, GHG emissions related to coal utilisation is a significant problem for Turkey, and these emissions must be lowered for a sustainable future. Under these

F.G. Üçtug
et al. / Journal of Environmental Management 132 (2014) 1e8

2

Post-combustion CCS

Fuel

CO2 emissions

On-site Steam Production Steam, O2

Fuel

Steam Reforming

Desulphurization

syngas (CO / CO2 / H2)

Compressor

Methanol Converter

Cooling and Distillation

Methanol

syngas recycle loop Fig. 1. Methanol synthesis from syngas.

circumstances, amongst many other options, conversion of CO2 into a usable product emerges as a promising solution, not only because of its environmental benefits, but also due to the economic gain. At this stage, methanol (CH3OH) appears as a viable product to be obtained via the CO2 in the flue gas, as suggested in the work of Sayah et al. (2010). 1.1. Methanol synthesis Methanol is a colourless, light, water-soluble liquid with a mild alcoholic odour. Globally, 70% of the methanol synthesised is used to produce formaldehyde, methyl-tert-butyl ether, and acetic acid. Chloromethanes, methylamines, methyl methacrylate and dimethylterephthalate are also produced from methanol. Paints resins, silicones, adhesives, antifreeze, plastics are some examples of methanol-based products (Cheng and Kung, 1994). Theoretically, methanol has high volumetric energy density compared to conventional batteries; hence, in the future direct methanol fuel cells are expected to replace conventional batteries in laptop computers and mobile phones, mainly due to being lighter in weight and
and Holmes, 2011). having extended operational lives (Üçtug Methanol can also be used in electricity generation plants as an alternative to natural gas. It provides the same flexibility as natural gas, including the ability to start, stop, accelerate and decelerate rapidly (Olah et al., 2006). The most common method of methanol production is methanol synthesis from syngas. Other methods include direct methane oxidation or biological selective conversion of methane (Olah et al., 2006; Crabtree, 1995). Methanol synthesis through syngas, which is a mixture of carbon monoxide (CO) and hydrogen (H2), involves the following chemical reactions (Fielder et al., 2003):

CO þ 2H2 4CH3 OH

(1)

CO2 þ 3H2 4CH3 OH þ H2 O

(2)

CO2 þ H2 4CO þ H2 O

(3)

Reaction (1) and (2) are exothermic and lead a reduction in volume. Conversely, reaction (3) is an endothermic reaction and is called “reverse water gas shift reaction” (RWGSR). RWGSR produces CO, which can be utilised to produce more methanol by reacting with hydrogen. It must be stated that the CO2 in the flue gas can be utilised in either reaction (2) or (3) or in both reactions simultaneously. However, in either case syngas would have to be produced separately; and this can be achieved by reforming or partial oxidation of coal, coke, natural gas or petroleum (Kochloefl, 1997). The flowchart of the methanol synthesis process can be seen in Fig.1. Although the chemical reactions associated with methanol production from syngas do not cause direct CO2 emission, 3.8 tonnes of CO2 per ton of methanol produced are indirectly emitted when the source of syngas is coal (Cifre and Badr, 2007). Some part of these emissions is caused by the on-site production of steam for the reforming process via fuel combustion, whereas the rest are offsite (indirect) emissions. It has been stated that indirect emissions account for approximately 3.5% of the total emissions (Methanex, 1996). Hence, by incorporating the data presented by Cifre and Badr (2007) with the data provided by Methanex Corporation (1996), it was concluded that approximately 3.67 tonnes of CO2 per ton of methanol are directly released and can be captured onsite during methanol production.

1.2. Carbon capture and sequestration (CCS) CO2 capture and sequestration (CCS) is attracting increasing attention as an option to reduce GHG emissions. CCS may, in fact, play a substantial role in the smooth and cost-effective transition to a sustainable, low-carbon energy future (Rubin et al., 2007; Turkenburg, 1997). CCS process consists of three steps in general: CO2 capture, CO2 transportation and CO2 sequestration. CO2 is captured at the source by using different techniques like absorption, adsorption, membrane-based separation or cryogenic separation (Steeneveldt et al., 2006). The captured CO2 needs to be in liquid or supercritical fluid state in order to be able to be

F.G. Üçtug
et al. / Journal of Environmental Management 132 (2014) 1e8

transported. The final step, sequestration, could be performed in three ways: geological sequestration, ocean sequestration and mineralisation. The main point of CCS systems is to keep the CO2 from being released to the atmosphere. CO2 capture can be realised by three different mechanisms: post-combustion capture, pre-combustion capture and oxy-fuel capture systems (Damen et al., 2006). Postcombustion capture technology can be used when the CO2 concentration of the exhaust gas is low, as in the case of hydrocarbonbased power generation or methanol production (Zanganeh et al., 2009). When high concentrations of CO2 are to be handled, as in the case of cement production, CCS systems become less costeffective. Absorption by using amine solutions such as aqueous monoethanolamine (MEA) as absorbent is the most common CO2 separation process. Post-combustion capture system produces high purity CO2, which is a feedstock for urea production or the food/ beverage industry. The fuel is reformed by oxygen and/or steam to form a mixture of H2 and CO2 in pre-combustion capture systems. After the separation process pure H2 is combusted in the power plant. Absorption, adsorption or membranes can be used to separate CO2 from the mixture (Blomen et al., 2009). Biomass and natural gas are more suitable for pre-combustion capture process. The carbon fuel is converted to carbon-free fuel with this system and chemical energy of carbon is transformed into chemical energy of hydrogen. The CO2 obtained via pre-combustion is usually more concentrated when compared to that of post-combustion. The most important obstacle for the widespread commercialisation of pre-combustion capture systems is the high investment cost (Olajire, 2010). Oxy-fuel combustion capture technology uses pure oxygen. Combustion products, CO2 and H2O, are separated by condensing water (Zanganeh et al., 2009). Combustion with pure oxygen leads to higher temperatures because nitrogen in air is the major heat sink (Kather and Scheffknecht, 2009). Cooled flue gases are used as heat sink in oxy-fuel combustion technology. The temperature is reduced to normal values during the recycle process of CO2-rich flue gas. The highest concentration of CO2 product is reached by using this system (Olajire, 2010). 1.3. Carbon trading Carbon trading is an approach used to control CO2 emissions by providing economic incentives for achieving emissions reductions. It is sometimes called cap and trade or carbon emissions trading, which refers to the trading of harmful emissions of six major GHGs e carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs) and sulphurhexafluoride (SF6). Carbon trading can be considered as a market-based instrument aimed at mitigating climate change. Currently there are several emissions trading schemes (ETSs) operating across the world. They differ in size, scope and design. Many ETSs have been proposed, but the only mandatory systems already operating are: The European Union Emissions Trading System, also called EUETS, the Regional Greenhouse Gas Initiative (USA), New Zealand Emissions Trading Scheme, Tokyo Metropolitan Trading Scheme (Japan) and the New South Wales Greenhouse Gas Abatement Scheme (Australia). Detailed information on each of these schemes can be found in the study of Perdan and Azapagic (2011). Cap and trade system is a commonly used approach in many of the above-mentioned trading schemes, including EUETS. The system operates by the allocation and trading of GHG emissions allowances to entities. One allowance gives the right to emit one tonne of CO2 equivalent. Absolute quantities limit (or cap) on CO2 emissions has been placed on each entity covered by the scheme.

3

Each entity then receives an allowance that is equal to its cap value. On a yearly basis, these entities are required to ensure that they have enough allowances to cover their emissions. Entities that keep their emissions below their cap value can sell their excess allowances at a price determined by supply and demand at that time. Entities that experience difficulties in remaining within their allowance limit have a choice between several options. They can take measures to reduce their emissions (such as investing in more effective technology or using a less carbon-containing energy source), buy extra allowances from the market or use a combination of the two. This flexibility ensures that emissions are reduced in the most cost-effective way (EC, 2008). 1.4. Motivation of the study In this study, we developed an optimisation-based economic model that determines whether installing a CCS unit, carbon trading or the combination of the two is a more beneficial option for a hypothetical methanol synthesis plant that utilises syngas as feed. The details of the model will be presented in the following section. Although the scope of this paper is limited to methanol synthesis via syngas, we believe that as long as the necessary model inputs such as CO2 emissions for the particular process of interest and the cost data for the chosen CCS method are available, our model can be applied to any industry. Hence, the methodology developed as a result of our study is as important as, if not more than, our findings. 2. Problem statement and mathematical model In this section, first the definition of the problem studied in this paper and the related literature are presented. Then, the mathematical formulation is explained in detail. Finally, information about the solution approach is provided. 2.1. Problem statement We studied a hypothetical methanol production facility that is considering installing a carbon capture and sequestration unit in order to ensure that the amount of CO2 emissions is within its allowable limits. This CCS unit, if decided to be built, has a certain capacity. We used cap and trade system, which requires each individual plant to emit CO2 less than or equal to its cap value. The planning horizon is T years, and the company decides how much of the capacity of the CCS unit to use each year. If the amount of capacity used in a given year is less than the allowable CO2 emissions cap for methanol production, then our hypothetical plant has to purchase credits from other entities to make up for the difference. On the other hand, if the amount is greater than the allowable cap, then our plant would be entitled to sell credits to other entities. Moreover, this CCS unit has a fixed capital investment cost that depends on its capacity and an annual operation and maintenance cost that depends on the amount of CO2 captured. The problem then becomes determining the capacity of the CCS unit to be built and the amount of CO2 captured by the CCS unit each year. In the literature, there are several papers that consider similar problems. Zhang et al. (2012) considered China’s power sector and analysed the effects of the implementation of three carbon mitigation policies. They developed an optimisation model and concluded that “Surplus-Punishment & Deficit-Award” carbon tax policy is the best option in increasing CO2 reduction effect and also reducing the total cost. Cristóbal et al. (2012a) studied a coal-fired power plant that considers installing four control devices in series, where it is possible to bypass one or more of them, before the CO2 discharge point. They developed a mixed-integer nonlinear

F.G. Üçtug
et al. / Journal of Environmental Management 132 (2014) 1e8

4

programming (MINLP) model in which they decided on which devices to use and the operating condition of each device. In Cristóbal et al. (2012b) a similar problem, where a set of pollution control retrofitting alternatives were considered to be installed in a cap and trade framework, is modelled as a MINLP. They demonstrated the capabilities of the model through a case study of an existing plant. 2.2. Mathematical formulation In this section, the mathematical formulation of the problem described above will be given. The decision that the company needs to make first is to decide on the CO2 capture capacity of the CCS unit. A low capacity would mean a relatively small initial investment as well as small operation and maintenance cost for the CCS unit, but would result in carbon trading becoming an expense item for the plant. A high CO2 capture capacity would require a larger investment and higher operation cost, but in that case, additional income could be generated via carbon trading. Hence, the problem can be viewed as an optimisation problem where the decision variable is the annual amount of CO2 captured by the CCS unit. The parameters and the decision variable of this optimisation problem are given below. Parameters: T: total useful life of the investment m : annual direct CO2 emission of the plant (ton/year) mi : CO2 emission allowance cap value in year i (ton) si : CO2 credit price in year i (V/ton) oi ðxi Þ : operation and maintenance cost of the CCS unit as a function of the CO2 captured in year i (V/ton) c(max i{xi}): capital cost of the CCS unit as a function of xi (V) r : interest rate

Since si, m and mi are all parameters, the second term evaluates to a constant in the equation above. Hence, neither m nor mi affects the optimal decision. They only affect the value of the objective function, i.e., the net present worth of the investment. The third term, oi(xi), is the operation and maintenance cost of the CCS unit that depends on the amount of CO2 captured. This cost can be any type of function. All these revenue and cost amounts are brought up to the current year at the interest rate of r. The last term in the objective function, c(max i{xi}), represents the capital investment cost. Our model might return different xi values for different years, the CCS unit must be capable of meeting the demand when xi is at its upper bound value. Therefore, the capital cost should be calculated for the maximum xi value, which will give the whole capacity of the CCS unit. Because of the technological restrictions, we assume that the decision on having a CCS unit is made at the beginning of the planning horizon when the methanol production facility is built. Hence, the capital investment cost is subtracted from the net present value of the revenue and cost in the initial year of the planning horizon. 2.2.2. Constraints There is a single set of constraint, which ensures that the amount of CO2 captured each year cannot exceed the amount of CO2 emitted by the plant, i.e. m. Then, we can write the optimisation model as a mixed integer nonlinear programming as follows:

Maximise subject to

zðxÞ 0  xi  m

for all i ¼ 0; .; T;

where continuous variables xi represent the amount of CO2 captured in year i, i ¼ 0,1,.,T. 3. Case study and scenarios

Decision variable: xi : amount of carbon captured by the CCS unit in year i (ton) z(x) : net present value of the profit (or negative cost) as a function of decision variable vector x (V)

2.2.1. Objective function The objective of this problem is to maximise the net present value of the profits that would result from pursuing a mix of the two optional actions described above over the total useful life of the investment, T.

zðxÞ ¼

T X i ¼ 0 ð1

1 i

þ rÞ

h i buy si COsell  oi ðxi Þ  cðmaxi fxi gÞ 2  si CO2

In this equation, COsell 2 is the amount of CO2 credits sold to other entities if the amount of CO2 captured (i.e., xi) exceeds the amount of CO2 that needs to be captured (i.e. m  mi ): buy COsell ¼ maxð0; xi  m þ mi Þ. CO2 is the amount of credits 2 bought from other entities if the amount captured is less than the amount of CO2 that needs to be captured: CObuy ¼ maxð0; m  mi  xi Þ. Therefore, the first two terms give us 2 the revenue (or negative cost) that will be obtained by carbon trading. Note that both COsell and CObuy cannot be positive at the 2 2 same time, which means the company either buys CO2 credits or sells them. Then, the first two terms in the objective function become: buy

si COsell 2  si CO2

¼ si xi  si ðm  mi Þ

(5)

The mathematical model that is developed in the previous section was applied to a hypothetical methanol production facility in Turkey. The values of the parameters used in the model are given below for this facility. 3.1. Parameters The first parameter to be determined is the annual direct CO2 emissions of the methanol production facility, m. As stated in Section 1.1, 3.67 tonnes of CO2 are directly emitted per ton of methanol produced via syngas. Therefore, the annual production capacity of the plant would determine the m value. Thus, for an annual methanol production capacity of n tonnes, m value would simply be calculated as:

m ¼ 3:67  n

(6)

Currently, there is no mass production of methanol in Turkey. Certain industries produce methanol as a by-product, however most of the market demand is supplied via imports. Methanol demand in Turkey has steadily increased since the beginning of 2000s, and future estimates point to values between 300,000 and 500,000 tonnes per year (Altinay, 2005). Under these circumstances, it was decided to keep the annual methanol production capacity (n) of our hypothetical plant at 300,000 tonnes per year, leading to an “m” value of 1,101,000 tonnes/year. To the best of our knowledge, CO2 emission cap value for methanol production is not available in the literature. In fact, emission cap values were found to differ from country to country, and within each country they differ from industry to industry.

F.G. Üçtug
et al. / Journal of Environmental Management 132 (2014) 1e8

5

Table 1 Parameters of the optimisation model. Parameter

Symbol

Unit

Value

Annual direct CO2 emission of the methanol plant Annual methanol production capacity of the plant Operational lifespan of the CCS unit Capital investment cost of the CCS unit Operation cost of the CCS unit in year i

m n T cðmaxi fxi gÞ o(i, xi)

Tonnes/year Tonnes/year Years V V

3.67 n 300,000 20 0:00003maxi fxi g2 þ 251:06maxi fxi g 50,000 þ 2x7/6 i

Therefore, assigning a constant m value would restrain our study and also lessen its reliability. Hence, we decided to construct scenarios in which we used different values for m, as presented in the next section. Similar to emission allowance caps, CO2 credit price is also difficult to accurately obtain. CO2 prices have displayed significant fluctuations in the past (Doty and Turner, 2009) and future estimations regarding these prices are various. At the beginning of phase 2 of EUETS, CO2 prices were expected to continuously rise, however recent trends show that they are in a period of decrease (CO2prices.eu, 2012). In fact, a previous estimation of V25/ton CO2 for late 2011 and early 2012 turned out to be highly inaccurate, with the actual values for that time period being approximately V5/ton CO2. Yet, there is no guarantee that CO2 prices will remain at their current levels throughout the useful life of our hypothetical CCS unit, T, which is taken as 20 years. The market prices in the future will be a function of many factors ranging from the national and international legislations setting the limits to reduce emissions, to the penetration of renewable energy technologies into the national energy systems. The more pressure the governments will lay on industry the higher the carbon credit prices should be expected in the future. Therefore, CO2 price estimation is a very challenging process, and it is left out of the scope of this particular study. Thus we decided to run our model at different values for the CO2 prices and obtain results. The reader is referred to Section 4 for a detailed analysis. The next term in the objective function is the capital cost. We obtained the cost and capacity limit data of five similar units (IPCC, 2005). Then, we fit a non-decreasing second order polynomial function to these data and obtained an investment cost function as follows:

cðmaxi fxi gÞ ¼ 0:00003maxi fxi g2 þ 251:06maxi fxi g

(7)

Note that this function passes through the origin, which means the investment cost equals to zero when the amount of CO2 captured equals zero. Moreover, this function is a non-decreasing concave function, meaning that as the capacity level increases, the investment cost increases with a decreasing speed. Finally, we need to calculate the operation and maintenance cost of the CCS unit as a function of the amount of CO2 captured in a given year, oi(xi). Cristóbal et al. (2012c) developed a mixed-integer nonlinear programming model that selects the best pollution control technologies for a coal-fired power plant. In their model they assume that the total operating and maintenance cost has two components: the fixed and variable operational costs. Kirschen and Strbac (2009) analyse the average cost function as the cost per unit of output of a facility. They stated that the average variable cost unavoidably rises as the processing amount increases, and explained this increase as follows: The fixed factors start constraining the process after some amount of production is achieved. For example, increasing the output more than a certain level might require maintaining the machines more frequently and in general implementing less-efficient procedures. As a result the variable processing cost turns out to be a convex function of the amount of

the output generated. Following Cristóbal et al. (2012c) and Kirschen and Strbac (2009) we decided to have an operation cost that has a fixed part and a variable part, and the variable cost is a convex function of the amount of CO2 captured. We also assumed that if the CCS unit is decided to be built, the fixed part of the operation cost is independent of the amount of CO2 captured, meaning that even if the unit is not being used in a given year, we need to pay for the fixed cost. Then, the operation and maintenance function takes the following form:

oi ðxi Þ ¼ a þ bðxi Þ

(8)

where a is the fixed part and b(xi) is a convex function of xi. For our case study, we assumed that the annual expenditure that needs to be made in order to maintain the CCS unit even if it does not function at all is approximately V50,000 and therefore the fixed cost, a, was taken as 50,000. The convex function was obtained as 7=6 bðxi Þ ¼ 2xi . A function of this particular form was deliberately chosen because when the capacity of the CCS unit used is much smaller than its maximum capacity, the operation cost is expected to increase almost linearly (Kirschen and Strbac, 2009). Note that this linear cost is equal to 2, which is calculated as a certain percentage of the largest possible CCS unit’s capital investment cost per unit capacity. In a study by Barranon (2006), this ratio was taken as 4% in the case of a methanol production plant. Since the CCS unit on its own is simpler to operate and maintain when compared to the entire methanol production plant, a smaller ratio such as 1% of the capital investment cost per unit capacity was accepted to be a more accurate representation of the variable part of the operation and maintenance cost. As calculated from Eq. (7) above, 1% of the maximum possible unit capital investment cost is approximately 2. Moreover, when the CCS unit is started to being used close to its capacity level, the operation cost will increase more rapidly because of the convexity of our function, which is in alignment with the process cost functions given in Kirschen and Strbac (2009). The list of all the parameters of our optimisation model can be found in Table 1. 3.2. Scenarios Some of the parameters that our model includes cannot be taken as known constant values. Therefore, we decided to create a number of scenarios and analyse the effect of these parameters on the optimal decision. The first parameter that we considered is the cap values on the CO2 emissions. In their study, Bleischwitz et al (2007) stated that according to the EUETS system, installations (entities) will be given initial allowances covering the first year that are equal to their annual CO2 emission levels, and gradually these allowances would be decreased. However, there is no data or comment on the nature of this gradual decrease. Thus, we decided to perform a sensitivity analysis by evaluating two different scenarios, one involving a 2% annual decrease in m, and the other involving a 3% annual decrease in m.

F.G. Üçtug
et al. / Journal of Environmental Management 132 (2014) 1e8

6 Table 2 Different scenarios analysed by the model. Scenario#

r

Dmi

Dsi

Scenario Scenario Scenario Scenario Scenario Scenario Scenario Scenario

5% 5% 5% 5% 10% 10% 10% 10%

2% 2% 3% 3% 2% 2% 3% 3%

2% 3% 2% 3% 2% 3% 2% 3%

1 2 3 4 5 6 7 8

Another parameter that affects our analysis is the interest rate, r. We created two different cases in which r ¼ 5% and r ¼ 10%. A main component that affects the optimal decision is the carbon credit price, si. As stated in Section 3.1, CO2 credit price is difficult to obtain accurately. It is known that low CO2 prices discourage entities from investing on CCS (Celebi and Graves, 2009). Therefore, we decided to run our analysis for different values of CO2 credit price ranging from V15to V40 for every scenario for the base year, which year 0. We also have two scenarios, in which si increases by 2% and 3% respectively, every year. As shown in Table 2, our analysis involves two different cases for the interest rate (r), two different cases for the rate of change in CO2 emission cap (-Dmi ), and two cases for CO2 credit price increase

z(x)

z(x)

200000000 150000000

y(x)

Euros

Euros

100000000 50000000 0 -50000000 15 20 25 30 35 40 45 -100000000 Base year CO2 credit price , s0 ( )

(a)

25

30

35

40

45

Base year CO2 credit price , s0 ( )

z(x)

y(x)

150000000 100000000 50000000 0 -50000000 15

20

25

30

35

40

y(x)

100000000 50000000 0

45

-50000000

-100000000

15 20 25 30 35 40 45

-100000000

(f)

Base year CO2 credit price , s0 ( ) z(x)

200000000 150000000

0

z(x) y(x)

50000000 Euros

50000000

Base year CO2 credit price, s0 ( )

100000000

y(x)

100000000 Euros

20

150000000

Euros

Euros

(e)

y(x)

z(x)

250000000 200000000

(b)

100000000 80000000 60000000 40000000 20000000 0 -20000000 15 -40000000 -60000000 -80000000

-50000000 15 20 25 30 35 40 45

0 15 20 25 30 35 40 45 -50000000

-100000000 -100000000

-150000000

(d)

Base year CO2 credit price, s0 ( )

z(x) 250000000 200000000 y(x) 150000000 100000000 50000000 0 -50000000 15 20 25 30 35 40 45 -100000000 -150000000

Base year CO2 credit price , s0 ( )

(g)

Base year CO2 credit price , s0 ( )

z(x)

150000000 y(x)

100000000 50000000

Euros

Euros

(c)

0 15

20

25

30

35

40

45

-50000000 -100000000

(h)

Base year CO2 credit price , s0 ( )

Fig. 2. Net present worth, z(x), and net benefits, y(x), associated with each scenario as a function of different base year CO2 credit prices. (a) Scenario 1, (b) Scenario 2, (c) Scenario 3, (d) Scenario 4, (e) Scenario 5, (f) Scenario 6, (g) Scenario 7, (h) Scenario 8.

F.G. Üçtug
et al. / Journal of Environmental Management 132 (2014) 1e8

7

Table 3 Annual amounts of CO2 captured associated with each scenario as a function of different base year CO2 credit prices. s0

15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

Annual amounts of CO2 captured (tons/year) Scen.#1

Scen.#2

Scen.#3

Scen.#4

Scen.#5

Scen.#6

Scen.#7

Scen.#8

0 0 0 0 0 0 0 0 0 0 0 0 37972 66133 113430 196380 365210 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000

0 0 0 0 0 0 0 0 0 0 45629 82984 149760 281040 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000

0 0 0 0 0 0 0 0 0 0 0 0 37972 66133 113430 196380 365210 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000

0 0 0 0 0 0 0 0 0 0 45629 82984 149760 281040 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000 1101000

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 49695 87054 155860 321040 1101000 1101000

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 35983 65991 121380 241410 1101000 1101000 1101000 1101000 1101000

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 49695 87054 155860 1101000 1101000 1101000

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 35983 65991 121380 241410 1101000 1101000 1101000 1101000 1101000

rate (Dsi). Thus, a total of 8 different scenarios were evaluated for different CO2 credit prices. In Table 2, the term -Dmi refers to the annual rate of decrease in the CO2 emission cap value. The results and discussions are given in the following section. 3.3. Solution method The mixed-integer nonlinear programming model explained in the previous section is implemented in the algebraic modelling system GAMS. Note that since the fixed cost is independent of the xi values, it does not directly affect the results of the optimisation problem. Therefore, we first solved the problem in its current form (including the fixed operation cost). Then, we calculated the objective function’s value when all xi values are zero (which means there is no capital investment or operations and maintenance cost but only a cost of buying CO2 emission allowances). If the objective value associated with the latter case is higher than the first one, we concluded that the optimal strategy is not building the CCS facility. If the converse is true, then we already have the optimal solution obtained by solving the mathematical programming formulation. 4. Results and discussion The first set of results to be presented is the net present worth of the profits and the net benefits. The net benefits, y(x), were determined by calculating the difference between the net present worth of the profits, z(x), and the total cost of purchasing CO2 credits throughout all 20 years without installing a CCS unit at all. Net benefit is an especially significant parameter because a positive net benefit value means that installing a CCS unit becomes a feasible action. A positive net benefit value does not necessarily mean that there is a net income as a result of the action, in certain cases it means that a particular action results in fewer deficits than its alternative. A net income would only be achieved in the case of a positive net present worth of investment value. The results for each scenario are given in Fig. 2.

The annual amount of CO2 captured by the CCS unit is also regarded to be an important finding of this study. Although our model allows the CCS unit to capture different amounts of CO2 every year, the results show that the optimal decision is to capture the same amount of CO2 every year during the 20 year lifespan of the CCS unit. Note that these findings depend on the values of the parameters whose effects were analysed in this case study. In Table 3, annual amounts of CO2 captured associated with each scenario as a function of different base year CO2 credit prices are given. When the effect of the rate of increase in the CO2 credits prices is analysed, it can be concluded that as the rate of increase gets larger in magnitude, achieving a positive net benefit can be realised at lower base year credit prices, s0. For instance, when Scenario 1 and Scenario 2 are compared, it can be seen that in Fig.2a the s0 value at which the y(x) value turns positive is higher than the corresponding s0 value in Fig.2b. The same effect can be observed in Scenarios 3 and 4, 5 and 6, and 7 and 8. When the scenarios that only differ in the rate of decrease in the annual CO2 emission caps are compared (see Scenarios 1 and 3, 2 and 4, 5 and 7, and 6 and 8), it was observed that the net benefits of such scenarios are identical as well as the optimal decisions on the amount of CO2 captured (see Table 3). In other words, net benefits were found to be only determined by the rate of interest and the CO2 credit prices. Note that from Section 2.2. It is already known that the cap values do not directly affect the optimal decision but only affect the objective function’s value. Our findings are in line with this theoretical result. However, when the effect of -Dmi value on the net present worth of the profits is analysed, we see that the net present worth values differ in all scenarios. For example; when Scenario 2 and Scenario 4 are compared, it can be seen that the net present worth becomes positive at s0 ¼ 34 for Scenario 2 while it becomes positive at s0 ¼ 37 for Scenario 4. The same observations can be made for other scenario pairs (see Fig. 2b and d). When the effect of the interest rate is analysed, the outcome is as follows: At the higher interest rates, the base year CO2 credit price that is required to achieve a positive net benefit value

8

F.G. Üçtug
et al. / Journal of Environmental Management 132 (2014) 1e8

becomes higher. Thus, it can be concluded that smaller rates of interest are more desirable, as expected. Another finding that can be obtained by analysing the results presented in Table 3 is that s0 values must be higher than the actual (real life) values so that CCS becomes a feasible choice. For instance, in Scenarios 5 and 7 an s0 value of V35 was found to be necessary for CCS to become feasible. As stated in Section 3.1, current CO2 credit prices are approximately V5/tonne, hence it can be concluded that real-life prices are actually far from encouraging investors into installing CCS systems. 5. Conclusion In this study, an optimisation model was established to decide on whether it is more economical for a hypothetical methanol production plant to invest on carbon capture and sequestration or to compensate for its CO2 emissions via carbon trading. The model allows us to determine a set of parameters which consequently result in CCS becoming a feasible action for the plant. The plant was assumed to use syngas from coal as raw material and the CO2 emissions were obtained as 3.67 tonnes per ton methanol produced. The annual production capacity of the plant was accepted as 300,000 tonnes. CO2 capture technology was selected as postcombustion CCS via amine scrubbing. The decision variable of the model is the annual amount of CO2 captured whereas the objective function was the net present worth of the profits. The nature of the operation and maintenance cost and the capital cost expressions caused the problem formulation to be in non-linear form. Algebraic modelling system, GAMS, was used to implement the optimisation model and to obtain the results. A total of eight different scenarios were evaluated, with two cases for the interest rate, r, two cases for the annual rate of decrease in the allowable CO2 emissions cap, LDm, and two cases for the annual rate of increase in carbon credit price, Dsi. The best case scenarios appeared to be Scenario 2 and Scenario 4, in which r ¼ 5% and Dsi ¼ 3%. These scenarios are regarded to be the best cases because in these scenarios, a positive net benefit value was achieved at the smallest value of the base year CO2 credit price when compared to the other scenarios. On the other hand Scenarios 5 and 7, which both involve an r value of 10% and an Dsi value of 2%, emerged as the worst case scenarios. Our findings show that the rate of interest is the most influential parameter, followed by the rate of increase in the carbon credit price. As indicated above, the methodology developed in this study is regarded to be the key point, rather than the analysis of methanol industry in Turkey, which is merely a case study so that our model can be applied. This model can be applied to any industry, any country, any type of CCS technology, as long as the required financial and technical data are available. References Altinay, B., 2005. General Information on Methanol. Specialisation thesis. Tobacco and Alcohol Markets Regulatory Authority, Turkey. Barranon, D.C.C., 2006. Methanol and Energy Production: Energy and Cost Analysis. M. Sc. thesis. Department of Applied Physics and Mechanical Engineering, Division of Energy Engineering, Lulea University of Technology, Sweden. Bleischwitz, R., Fuhrmann, k., Huchler, E., 2007. The sustainability impact of the EU emissions trading system on the European industry. In: Bruges European Economic Policy Briefings, vol. 17. College of Europe, Belgium. Blomen, E., Hendriks, C., Neele, F., 2009. Capture technologies: improvements and promising developments. Energy Proced. 1, 1505e1512.


lu, I., 2012. Present situation and potential role of Capik, M., Yılmaz, A.O., Çavus¸og renewable energy in Turkey. Renew. Energy 46, 1e13. Celebi, M., Graves, F., 2009. Volatile CO2 Prices Discourage CCS Investment. The Brattle Group, USA. Cheng, W.H., Kung, H.H., 1994. Methanol Production and Use. Marcel Dekker, New York. Chicco, G., Stephenson, P.M., 2012. Effectiveness of setting cumulative carbon dioxide emissions reduction targets. Energy 42, 19e31. Cifre, P.G., Badr, O., 2007. Renewable hydrogen utilisation for the production of methanol. Energy Convers. Manage. 48, 519e527. CO2prices.eu, 2012. http://www.co2prices.eu/. retrieved on July 26, 2012. Crabtree, R.H., 1995. Aspects of methane chemistry. Chem. Rev. 95, 987e1004. Cristóbal, J., Guillén-Gosálbez, G., Jiménez, L., Irabien, A., 2012a. Optimisation of global and local pollution control in electricity production from coal burning. Appl. Energy 92, 369e378. Cristóbal, J., Guillén-Gosálbez, G., Jiménez, L., Irabien, A., 2012b. MINLP model for optimising electricity production from coal-fired power plants. Energy Policy 51, 493e501. Cristóbal, J., Guillén-Gosálbez, G., Jiménez, L., Irabien, A., 2012c. Multi-objective optimisation of coal-fired electricity production with CO2 capture. Appl. Energy 98, 266e272. Damen, K., Faaij, A., Turkenburg, W., 2006. Health, safety and environmental risks of underground CO2 sequestration e overview of mechanism and current knowledge. Clim. Change 74, 289e318. Doty, S., Turner, W.C., 2009. Energy Management Handbook, seventh ed. The Fairmont Press Inc., USA. EC, 2008. Eu Action against Climate Change: the EU Emissions Trading Scheme. Office for Official Publications of the European Communities, Luxembourg. Fielder, E., Grossmann, G., Burkhard Kersebohm, D., Weiss, G., Witte, C., 2003. Methanol. In: Ullmann’s Encyclopedia of Industrial Chemistry, sixth ed., vol. 21. Wiley-VCH, Weinheim, pp. 611e635. GAMS Home Page, http://www.gams.com/, retrieved on August 28, 2012. IPCC, 2005. IPCC Special Report on Carbon Dioxide Capture and Storage. Cambridge University Press, Cambridge. IPCC Fourth Assessment Report: Climate Change 2007 (AR4). http://www.ipcc.ch/ publications_and_data/publications_and_data_reports.shtml#1, retrieved on July 9th, 2012. Kather, A., Scheffknecht, G., 2009. The oxycoal process with cryogenic oxygen supply. Naturwissenchaften 96-9, 993e1010. Kirschen, D.S., Strbac, G., 2009. Fundementals of Power System Economics (Chapter 2). John Wiley & Sons Inc., West Sussex, England, p. 30. Kochloefl, K., 1997. Handbook of Heterogeneous Catalysis, vol. 4. Wiley-VCH, Weinheim. Methanex Corporation Climate Change Voluntary Challenge & Registry Program Action Plan September 1996. http://www.ghgregistries.ca/registry/out/C969METHANEX-W52.PDF, retrieved on May 17th, 2013. Olah, G.A., Goeppert, A., Prakash, G.K.S., 2006. Beyond Oil and Gas: the Methanol Economy. Wiley-VCH, Weinheim. Olajire, A.A., 2010. CO2 capture and separation technologies for end-of-pipe applications e a review. Energy 35, 2610e2628. Perdan, S., Azapagic, A., 2011. Carbon trading: current schemes and future developments. Energy Policy 39, 6040e6054. Roddy, J.D., 2012. A syngas network for reducing industrial carbon footprint and energy use. Appl. Therm. Eng. http://dx.doi.org/10.1016/j.applthermaleng.2012. 02.032. Rubin, E.S., Chen, C., Rao, A.B., 2007. Cost and performance of fossil fuel power plants with CO2 capture and sequestration. Energy Policy 35, 4444e4454. Sayah, A.K., Hosseinabadi, S., Farazar, M., 2010. CO2 abatement by methanol production from flue-gas in methanol plant. World Acad. Sci. Eng. Technol. 69, 90e93. Soltanieh, M., Azar, K.M., Saber, M., 2012. Development of a zero emission integrated system for co-production of electricity and methanol through renewable hydrogen and CO2 capture. Int. J. Greenhouse Gas Control 7, 145e 152. Steeneveldt, R., Berger, B., Torp, T.A., 2006. CO2 capture and sequestration e closing the knowing-doing gap. Chem. Eng. Res. Des. 84, 739e763. Turkenburg, W.C., 1997. Sustainable development, climate change and carbon dioxide removal (CDR). Energy Convers. Manage. 38, S3eS12.
, F.G., Holmes, S.M., 2011. Characterization and fuel cell performance analysis Üçtug of polyvinylalcoholemordenite mixed-matrix membranes for direct methanol fuel cell use. Electrochim. Acta 56, 8446e8456. Zanganeh, K.E., Shafeen, A., Salvador, C., 2009. CO2 capture and development of an advanced pilot-scale cryogenic separation and compression unit. Energy Proced. 1, 247e252. Zhang, D., Ma, L., Liu, P., Zhang, L., Li, Z., 2012. A multi-period superstructure optimisation model for the optimal planning of China’s power sector considering carbon dioxide mitigation. Energy Policy 41, 173e183.