Design of a crude distillation unit and its preheat train. By Božidar Aničid and Tibor Kuna

Transcription

1 Design of a crude distillation unit and its preheat train By Božidar Aničid and Tibor Kuna

2 Contents 1 Introduction Process description Product and feedstock prices Model results Economic performance Sensitivity analysis Summer winter scenarios Future work Conclusion Acknoledgements Bibliography APPENDIX A... 15

3 Abstract Our simulated model of a crude distilation unit, its preheat train and natural gas burning furnace gives clear insight into what kind of profits can be made if the heat integration is done correctly and feed composition selected wisely. Three different feed stocks namely Iraq heavy crude (Basrah), Iran heavy crude (Iranhy) and North sea Brent Blend (BrentBL) are used. A feed composition such that every feed stock represented one third in weight percent, was selected to be the base case in the preheat train optimization. In second part of this work a feed composition optimization is also performed. The preheat train consists of heat exchangers that incrementally heat up the crude feed using hot streams from the atmospheric distillation unit pumparounds and natural gas combustion flue gases. The internal rate of return for this similated unit capable of distiling ton/day of crude without integration was calculated to be 59%, and for an integrated process 76%. The optimal feed composition in the year 2009 was - BrentBL : Iranhy : Basrah = 0 : 0,4 : 0,6. 1 Introduction The advancement of computer power and dedicated software packages has given us the ability to integrate and optimize industrial processes like never before and thus bringing water and energy usage to a minimal, because water and energy are no longer seen as commodities, but as valuable assets. Companies must plan and execute their processes so that they are within ecological standards and provide reasonable profit not only today but also in the near future. Most of the times this is easier said than done. They can be challenging to optimize, but provide great rewards if they are brought up to date correctly. A lot of refineries running today were built in the 80's or late 70's when energy was cheap, and when the investors did not devote much attention to the costs of energy. Because of that leading oil companies carried out rationalization and suggested energy-saving programs. These programs consist of the following actions: (1) (2). Continuous monitoring of energy costs, identifying the places of irrational energy consumption and preparing the energy saving project modernization of equipment and introduction of computer management, reconstruction of equipment and intensification of the maintenance process. The prices of products in the petrochemical world are quite volatile because of the constantly shifting crude oil market, rising global demand, and the geopolitical situation. Therefore it is desirable that refineries are capable of operating with different crude oil compositions. 1

4 2 Process description The described process is for the base case where the crude mix at the begining consists one thrid of the Iranian heavy crude (IRANHY), one third North sea Brent blend (BrentBL), and one third Iraqi heavy crude (BASRAH) by weight percent. All of these assays can be found in the Aspen Plus library. Their specific properties are described with the use of a wast array pseudo components. In order to simulate effect of different feed composition on annual profit, different feedstock combinations have been used. The API gravity for the different feedstock combinations can be seen in table 1. Table 1-API gravity for different feedstock combinations. Basrah 0 0,8 0,6 0,4 0,2 1 0,8 0,6 0,4 0,2 0 0,6 0,4 0,2 0 0,4 0,2 0 0,2 0 0 Iranhy 1 0,2 0,4 0,6 0, ,2 0,4 0,6 0,8 0 0,2 0,4 0,6 0 0,2 0,4 0 0,2 0 BrentBL ,2 0,2 0,2 0,2 0,2 0,4 0,4 0,4 0,4 0,6 0,6 0,6 0,8 0,8 1 API Gravity 33,6 32,6 32,1 31,5 33,7 34,5 33,9 33,4 32,8 32,3 35,2 34,7 34,1 33,6 36,0 35,4 34,9 36,7 36,2 36,2 37,4 Preflash tower The crude oil that is entering the preflash tower (FEED) is at 20 C and is firstly preheated up to 154 C after passing through the preheat train which consists of two heat exchangers (B12, B11) which recover heat from the flue gases that come from the natural gas combustion. The Preflash tower (PF) has 10 equilibrium stages, and in this simulation the preheated crude enters in the bottom (stage 10) at a rate of ton/day. The crude is then directed towards the PF furnace which was replaced with a heat exchanger (B16) for the purpose of the simulation. This heat exchanger (B16) recovers heat from the reactor (B14). The crude then reaches its desired temperature of 232 C at 4 bar. Steam (STEAM-1) is also introduced at the bottom of the column. The base case furnace heat duty is 52 MW, while the condenser duty is 26 MW. The top stream (LIGHT) which consists of the lightest components, is drawn at the top of the column and purged. Although in real life this wouldn't be economical since you could use it as a feedstock for the furnaces or sell it as liquefied petroleum gas (LPG), but for this simulation none of the previously mentioned uses could be employed since the stream is too complicated to be used as a feedstock (because of pseudo components), and it has an API of 89 which is not high enough to be classified as LPG (3). In order to obtain necessary heat in the furnaces natural gas is used. It's composition is much simpler and currently it's price is lower compared to LPG price (4). For the main distillate light naphta (LNAPHTA) in the PF tower a design specification was set in Aspen Plus so that it always has an API of 75 (3). The PF is represented by the blue color in figure 1. Atmospheric distillation unit The heavier components of the crude oil that leave at the bottom of the preflash tower (ADU-1), then go through the second preheat train which consist of two heat exchangers (B2, B10) that increase its temperature from 215 C to 273 C. These two heat exchangers recover heat from the ADU pumparounds and flue gas from natural gas combustion in the reactor (B14). Finally the 2

5 preheated crude enters the furnace which gives the crude its final temperature of 380 C. Just as in the preflash unit, the furnace in the ADU is replaced with a heat exchanger (B3) for the simulation purposes. The furnace vaporizes a major portion of the crude and feeds this vapor-liquid mixture into the atmospheric distillation tower. The overflash in Aspen Plus is set to 3%. The ADU has two pumparounds, three sidestrippers and just like in the PF, its own steam supply (S5) at the bottom. The steam serves to strip any residue and prevent excessive thermal cracking of crude due to high temperatures (2). The pumparounds are used to reduce the vapor flow in the column and allow for heat recovery. The first pumparound is located from stage 8 to 6 and the second from stage 14 to 13. They recover 15,7 and 22,2 MW respectively. The three sidestrippers are present in order to draw various side streams at different locations, in our case from stage 18, 13 and 6. The draw locations represent the temperature range of the liquid products that we can collect from the given draw location. The side products that are drawn from the three sidestrippers are kerosene, diesel and atmospheric gas oil. The sidestripper are defined with a design specification so that they always draw a mixture with a specified API no matter what the feed composition. The first sidestripper draws a fraction with API of 57 (KEROSENE), the second is set to API of 40 (DIESEL) and the last with an API of 27 (AGO) (3). The base case furnace duty for the ADU is 121 MW, while the condenser duty is 96 MW. The ADU is represented by the yellow color in figure 1. Stoichiometric reactor and steam generation The natural gas flow rate (NG) towards the reactor (B14) is such, that both furnace duties were covered, so for the base case 173 MW. The reason why reactor is used for simulation of natural gas combustion is easier simulation of mentioned process. As stated before, both furnaces were replaced by two heat exchangers (for easier process simulation), which transfer the required heat duty from the natural gas combustion reaction to the crude. Air, which is also needed in the combustion, is supplied through the stream (AIR). The reactor is represented by the green color in fig. 1. The steam needed for the sidestrippers (STM-1, STM-2, STM-3) is generated by heating the water stream (WTR) with the recovered heat from the flue gases. The heat recovery is organized so that water coming in (WTR) is brought up to the specified pressure of 4 bar by the pump (B9), then it reaches boiling temperature in the heat exchanger (B8). The vaporization occurs in heat exchanger (B4). The steam then flows towards the last heat exchanger (B7) which gives its final temperature of 220 C. This generated steam (S9) is then combined with an auxiliary steam (CU-STEAM) in the mixer (B5). Finally the steam is sent towards a splitter (B6) which splits it into six streams (S13, STM-1, STM-2, STM-3, S5, STEAM-1), and then distributes them to all the units, except for (S13) which is purged. Reason why S13 is present is easier convergence. 3

6 Figure 1 - The process flow diagram is broken down into four main sections. The Blue section represents the preflash section, yellow color represents the atmospheric distillation section, the red is for the preheat trains, and green for the natural gas combustion and steam generation/distribution section. 2.1 Product and feedstock prices ADU process unit is one of the first units in the overall refinery process, meaning that the outputs of the ADU unit undergo further processing and blending, before they take their final form. For example, kerosene product stream, from the ADU, undergoes several more processes before final product, kerosene, is obtained. Prices of semi products (outputs of ADU unit) are hardly available, that is why the prices of final products, together with some assumption, must be taken into consideration for estimation of mentioned semi product s prices. Prices of some final products (kerosene, diesel, gasoline) are available on the EIA (5). Furthermore, very useful information is 4

7 Price ($/CUM) Price ($/CUM) obtained from the mentioned web page; the refinery process contributes from 10 to 15 % of the final product price (depending on the product) (5). That means that the price of a semi product is approximately 10 to 15 % smaller than the price of the final product. For example, it is assumed that the price of the diesel product stream (from the ADU) is 15 % smaller than the price of the final diesel product, and the kerosene product stream is 10 % smaller than the price of the final kerosene product (5). Prices of Ago semi product and ADU residue were unavailable. Therefore, in order to calculate needed prices, the price ratio between after mentioned prices of semi products and before calculated semi products prices (kerosene, diesel, gasoline) is taken into account (6). Finally, estimated prices of all semi products are represented at the Figure 2. Data used for obtaining graphs on figure 2, is given in the appendix A1. At the other hand the prices of feedstock, including prices of crude oils and natural gas, are represented at the figure 3. The same price variation trends can be observed for both, products and feedstock. It is assumed that the prices of raw material, including steam and cooling water, are constant. The price of steam is 0,055 $/kg, and the price of cooling water is 0,006 $/MJ. Since, for obtaining necessary temperature, natural gas is combusted, also the emission of CO 2 must be taken into consideration. It is assumed that the value of carbon tax is constant, and equals to 20 $/t of CO 2 emitted (7) diesel kerosene gasoline AGO ADU-residue Year Figure 2: Prices of ADU products during the investigated time period OPEC EU Brent Natural gas* Year Figure 3: Prices of the feedstock during the investigated time period. 5

8 3 Model results The results of the proposed crude distillation unit and its preheat train is presented in the following subsections. Mass flows of the main products and feedstock s for the base case are presented in table 2. The consumption of natural gas is presented for the first year of operating. Since the fouling has been taken into account, the amount of natural gas is increased by 5 % per year. This increase on gas consumption is taken from reference (8). Simulation results for the part where feed composition is varied are presented in table 3. These results are used for further calculation of economic performance. Table 2: Simulation results for base case Crude feed combination Basrah Iranhy BrentBL Feed Heavy naphta Light naphta Volumetric flow (m 3 /hr) Kerosene Diesel Ago ADU residue Natural gas 0,333 0,333 0, ,0 200,5 79,8 259,7 275,6 109,9 247, ,3 Table 3: Simulation results for different feed composition Crude feed combinations Basrah Iranhy BrentBL Feed Heavy naphta Light naphta Mass flow (kg/hr) Kerosene Diesel Ago ADU residue Natural gas ,8 0, ,6 0, ,4 0, ,2 0, ,8 0 0, ,6 0,2 0, ,4 0,4 0, ,2 0,6 0, ,8 0, ,6 0 0, ,4 0,2 0, ,2 0,4 0, ,6 0, ,4 0 0, ,2 0,2 0, ,4 0, ,2 0 0, ,2 0,

9 The overall energy balance is represented using Sankey diagram, and can be seen at figure 4. Presented diagram is for integrated process. It can be seen that flue gas heat is used for preheating of feedstock (B11, B12) and for steam production (green part on figure 1). The difference between integrated and non-integrated process is in fact usage of flue gases. In contrary to integrated process, in non-integrated process, flue gases are used only for preheat of feedstock. Figure 4- Sankey diagram of proposed CDU. 3.1 Economic performance The method of net present values (NPV) is used in order to determine the economic efficiency and validity of the overall process. The NPV is a commonly-used method for the calculation of economic feasibility. The value of the capital cost is one of the variables that must be calculated in order to determine NPV. Several parameters and indexes are used for the calculation of capital cost (9). The e main equation used for estimating capital (investment) costs is C n C S n S / ( ) f,where C is the 0 0 cost of a process unit, n is the number of equally-sized units, C 0 is the cost of a reference unit, S is the capacity of a process unit, S 0 is the capacity of a reference process unit, e is the cost scaling exponent for different numbers of equally-sized units, and f is the cost- scaling factor (10). Costs of reference units are taken from 2003, meaning that no further adjusting, using CEPCI indexes is necessary. Prices of the atmospheric distillation unit and complete site utilities (tanks, sewage treatments, etc.) are represented in the table 4. In order to calculate annual profit, the prices of raw material and feedstock are necessary. Those prices are mentioned in the previous chapter. Annual profit is calculated using following equation: A p = P p - P c - P ng O t M t, where, P p is the price of the total amount products, P c is the price of the total amount of crude oil used, P ng is the price of the total amount of natural gas used, O t are total operating costs, and M t is the maintenance cost. Operating costs are associated with usage of steam, cooling water, etc. The value of the maintenance costs varies between five and six percent of the 7

10 Profit (M$) capital costs (11). In our work it was estimated that the maintenance cost would amount to 5.5% of capital costs, resulting in 36,2 M$ per year. Table 4: Calculation of capital costs. Unit Scaling parameter S S 0 C 0 (M$) n f e C (M$) Atmospheric distillation unit Bbl/day , ,67 0,9 150,5 Site facilities Bbl/day ,8 33 0,67 0,9 506,7 Total cost 657,2 Finally, the last parameter needed for the calculation of NPV is the annual discount rate. The value of annual discount rate that is taken into consideration is 10%. Furthermore, the 30% tax rate is also taken into account. The calculated annual profits, for both processes, with and without heat integration, are represented at the figure ,00 700,00 500,00 300,00 100,00-100,00-300, with heat integration without heat integration -500,00-700,00 Year Figure 5: Calculated annual profits Calculated annual profits are higher for integrated process for about 1M$ per year. That proves efficiency of suggested heat integration. As it can be seen, annual profits are considerably smaller in years from 2009 to 2013 (in 2011 even negative) compared to first five years. That is a result of global economic crisis, and as it can be seen in the figures 2 and 3 the price of crude oils and products strongly decreased when crisis had started. As mentioned before, prices of crude oil and products are strongly connected (easily observed at figures 2 and 3), but on the other hand the price recovery of the crude oil was faster, compared to the price recovery of the refinery products. That is why, in the years after main impact of crisis, the annual profit strongly decreased (in 2011 even negative) in 8

11 NPV (M$) comparison to the profit obtained in the years before global crisis ( ). It is also very important that with decrease of profits, the heat integration becomes more obvious, because it contributes higher percentage of profit. Calculated annual profits are used for calculation of NPV. Results are represented at the figure With heat integration Without heat integration YEAR Figure 6: Net present value The NPV of integrated process is higher for about 4 M$ after the 10 years long period (2013). It is obvious that NPV achieved approximately constant value after year That is because the annual profit strongly decreased in 2009, resulting in negligible change in NPV value. Calculated internal rate of return for integrated process is 76%, and 59% for non-integrated process. Even though calculated NPV difference is only 4M$, this value has a strong influence on IRR, since IRR for integrated process is 17% higher. This fact, once again, proves efficiency of performed heat integration. 3.2 Sensitivity analysis In order to determine the effect of different values of independent variables, such are capital costs, price of products and reactants, on net present value, the sensitivity analysis is performed. The value of capital cost is varied ± 20% and ± 10 %. The variation influence is measured using calculation of internal rate of return. Furthermore, the prices of refinery products and crude oil are varied in the same manner. Results of the sensitivity analysis are represented on Figures 7, 8 and 9. When the product price change was 10%, obtained NPV was negative, meaning that calculation of IRR was not possible (it is smaller than proposed annual discount rate). Same occurs as the crude oil price increased by 10%. That is why further change (up to 20%) is omitted. It can be seen that the change in crude oil price and refinery product s prices has a strong impact on IRR, meaning that refinery process is very sensitive to products and feedstock price change. The feed composition is another variable that has been changed in order to determine its influence on economic performance. 9

12 IRR (%) IRR (%) IRR (%) with heat integration without heat integration with heat integration without heat integration 20-5% 0% 5% 10% 15% 20% Change in products price Figure 7: Effect of crude oil price change on IRR % -15% -10% -5% 0% 5% Crude oil price change Figure 8: Effect of change in products price on IRR with heat integration without heat integration % -10% 0% 10% 20% Capital (investment) cost change Figure 9: Effect of capital cost change on IRR The feed composition is another variable that has been changed in order to determine its influence on economic performance. Mass flows (of all products and feed stream) needed for calculations of annual profits are already represented in Table 3. Calculated annual profits, for different feed composition, are represented at the ternary diagram (Figure 10). Spots with highest annual profits are represented with red color, while dark blue represents lowest annual profit. Highest annual profit is obtained for feed composition BrentBL:Iranhy:Basrah = 0:0,4:0,6. The lowest annual profit (even negative) is obtained when pure BrentBL crude oil is used. Nonetheless, there are also some other spots with negative revenues. That is why optimal feed composition is crucial for positive economic performance of refinery process. This was reasonable to expect, due to the fact that composition of crude oils varies strongly, and crudes from which more expensive products are produces, are more economically favorable. In our case that are Iranhy and Basrah. Calculations of after mentioned annual profits are given in the appendix A2. It is well known that majority of crude oil fields are placed in countries with unstable political and social situation. Because of that, the crude oil supply from these countries is unsustainable, meaning that refineries from the other countries, which use crudes from mentioned areas, have to consider 10

13 alternative oil, when political or social situation break crude oil distribution. Alternative choice should be considered very carefully, because, as it can be seen from figure 10, same feed compositions can lead to the negative revenues. Figure 10: Variation of annual profits (2009) due to the change in feed composition Summer winter scenarios Crude oil distillation would be also evaluated through two summer-winter scenarios. In first such scenario, variation of diesel price was taken into consideration. It is well known that the price of diesel increases during winters, because the usage of heating oil strongly increases in winter mounts (12). Heating oil and diesel are produced is similar way, from same crude oil fraction. That is why the increase of heating oil consumption is connected with increase in diesel price. It is reasonable to expect that refinery should operate with feed composition that would give as much as possible diesel in the winter. Two extreme prices (winter and summer one) for the year 2010 are taken into consideration. The summer price of diesel is 812,8 $/m 3 and winter price is 981,6 $/m 3. It is assumed that one half of annual production is sold for summer price, and another half for winter price. The half year revenues are than compared, for different feed compositions. Results are graphically represented on Figure 11, while detailed calculations can be found in the appendix. Winter half year revenues are represented on the left hand side, while summer half year revenues are represented on the right. Same as in figure 10, red spots represents highest annual profits, while blue spots represent lowest annual profit. It is easily observed that there are huge differences between mentioned half year revenues, for all feed compositions. That once again shows how sensitive profit is on price change of one or more products. It can be also seen that majority of summer annual profits are negative, meaning that optimal feed composition is crucial for obtaining positive revenues. Highest winter revenue is obtained when pure Basrah is used, while highest summer profit is obtained for following feed composition : BrentBL : Iranhy : Basrah = 0 : 0,4 : 0,6. It is obvious that optimal feed composition changes when the price of product changes. That means that flexibility of refinery process, in a way that feed composition can be varied, is essential for obtaining highest annual profit. 11

14 Figure 11 - Variation of half year revenues due to the change in feed composition. Second summer winter scenario is associated with yearly temperature variation, and it is considered for base case (base feed composition). It is well known that in continental climate areas, air temperature varies by more than 40 C during the year. For example, winter temperature is most the time negative, while summer temperature is approximately 30 C. Different ambient temperature contribute to different feed temperature and heat losses, meaning that greater amount of natural gas is needed during winters, for satisfying heat needs. It is calculated that consumption of natural gas is 1000 kg/h greater, when feed temperature is -10 C compared to the 20 C feed temperature. The difference can have strong effect on annual revenue, since it can increase process annual cost by 1 to 1,5 M$. 4 Future work The overall refinery process is very complex, and CDU only a first part of mentioned process. In order to obtain better heat integration, the overall refinery process must be simulated more thoroughly. For example, also VDU, cracking units, hydrogenation process unit, etc. have to be incorporated for integration of mentioned units with CDU. Furthermore, the integration of some other process, in which energy is released, with CDU can be considered. Since CDU is energy consuming, such integration can have a very strong positive influence on economic efficiency. One of such process is methanol production, and instead of using the released heat for electricity production, it can be used within refinery process. Methanol process can be useful also from other point of view, since methanol is used as additive in gasoline. A mass integration of some streams (LIGHT) has not been looked at with great detail in this work, since it would increase the complexity of the simulation by great degree. In the future work the mentioned stream should be integrated, thus providing a feedstock for furnace or a new product (LPG). During our work in ASPENPlus we had difficulties using optimization function. That is why presented results are heuristically obtained. In order to increase overall efficiency, optimization of overall process must be performed (using ASPEN or some other program). 12

15 5 Conclusion The economical performance of the simulated refinery in this work was looked at from different points of view. The investigation can be placed into three main topics: a) The impact of the feedstock composition on the economical performance for the years 2009 and 2010 was studied. For this part we have chosen that the plant was integrated. The results are shown graphically with the help of ternary contour diagrams. b) The sensitivity analysis for the capital costs, product prices and feedstock price was also carried out. The variation influence was measured using calculation of internal rate of return. c) The impact of different variables (fouling, seasonal fluctuation in feed stream temperature, seasonal fluctuation in product prices, heat integration) was examined for a base case. This study concludes that the feedstock composition is of great importance, because on one hand we can achieve a great profit if the composition is right (Basrah:Iranhy:BrentBL=0,6:0,4:0), on the other hand (Basrah:Iranhy:BrentBL=0:0:1) we can also achieve big loss if we aren't careful when selecting the crude oil. The sensitivity analysis shows that, annual profit is very sensitive on feedstock or predicts price change. For example, the 10% increase in crude oil price, would contribute to negative NPV after the 10 years long period. As it was expected lower feedstock temperature contribute to a grater consumption of natural gas, thus lowering the revenue. The same could be said for fouling which also raised the consumption of NG. To sum up, it is shown that crude oil distillation can be very profitable process when the right feed composition is chosen. Furthermore, in order to increase economic performance an additional work (overall process simulation, optimization, etc.) should be done. 6 Acknoledgements We would like to thank Miloš Bogataj from the laboratory for process systems engineering and sustainable development at the faculty for chemistry and chemical technology Maribor, Slovenia for his mentorship and helpful insights in the preparation of this work. 13

16 7 Bibliography 1. Ocic, Ozren. Oil Refineries in the 21st Century Ai-Fu Chang, Kiran Pashikanti, Y. A. Liu. Refinery engineering - Integrated Process Modeling and Optimization. s.l. : Wiley-Vch, Exxonmobil. [Online] 4. Indexmundi. [Online] [Cited: 03 22, 2014.] 5. EIA. [Online] 6. Optimization of crude distillation system using aspen plus: Effect of binary feed selection on grassroot design. R.K. More, V.K. Bulasara, R. Uppaluri, V. R. Banjara. 2010, Chem. Eng. Res. Des., pp Slovenije, Carinska uprava Republike. Emisija CO2. [Online] [Cited: 9 1, 2013.] 8. Ernst Worrell, Christina Galitsky. Energy Efficiency Improvement and Cost Saving Opportunities For Petoleum Refineries. s.l. : Berkley National Laboratory, John D. Jones, P. E. Feasibility Study for a Petroleum Refinery for The Jicarilla Apache Tribe. Pheonix : s.n., Comparison of coal IGCC with and withour CO2 capture and storage. Martelli E., Kreutz T. 14, s.l. : Energy Procedia, 2009, Vol Pintarič, Zorka Novak. Razvoj produktov in procesov. Maribor : FKKT MB, Wikipedia. [Online] [Cited: 04 10, 2014.] Lasse R. Clausena, Niels Houbak b. Technoeconomic analysis of a methanol plant based on gasification of biomass and electrolysis of water

17 8 APPENDIX A Table A 1: Prices of various commodities. Year Opec ($/m 3 ) BrentBL ($/m 3 Diesel ($/m 3 ) Kerosene ($/m 3 ) Gasoline ($/m 3 ) AGO ($/m 3 ) ADU residue ($/m 3 ) Natural gas ($/m 3 ) ,10 320,62 406,43 297,96 425,44 325,29 190,30 0, ,36 457,30 539,36 426,98 582,99 445,75 260,77 0, ,85 546,04 607,39 488,32 660,89 505,31 295,62 0, ,89 607,05 647,81 537,14 714,38 546,21 319,54 0, ,49 812,36 853,94 747,40 964,62 737,55 431,48 0, ,68 517,38 553,95 427,78 592,49 453,02 265,02 0, ,03 667,13 671,84 533,28 727,01 555,87 325,19 0, ,51 932,36 862,25 734,96 962,52 735,94 430,54 0, ,19 935,46 890,99 747,48 987,62 755,13 441,76 0, ,19 909,73 880,66 727,70 969,70 741,43 433,75 0,13 Table A 2: Profit as a function of crude feed composition for the years 2009 and 2010, and for the winter/summer extreme. The numbers in the red color are for the biggest loss, and the green for the biggest profit. Weight fraction Basrah Iranhy BrentBL 2009 Profit(M$/a) Winter half revenue (2010) Summer half revenue (2010) ,2 87,0-67,5 0,8 0, ,3 283,4 63,2 0,6 0, ,7 271,9 99,1 0,4 0, ,4 173,0 6,3 0,2 0, ,1 128,9-29, ,0 287,6 64,1 0,8 0 0,2 346,0 226,1-15,4 0,6 0,2 0,2 245,4 143,5-68,9 0,4 0,4 0,2 239,2 124,0-57,7 0,2 0,6 0,2 180,6 74,5-89,1 0 0,8 0,2 92,1 9,0-144,5 0,6 0 0,4 323,1 198,3-23,2 0,4 0,2 0,4 285,5 155,6-31,7 0,2 0,4 0,4 214,7 94,6-67,4 0 0,6 0,4 87,9 3,0-148,7 0,4 0 0,6 276,2 161,4-56,6 0,2 0,2 0,6 154,8 69,2-130,0 0 0,4 0,6 97,6 28,5-168,5 0,2 0 0,8 229,2 123,7-89,2 0 0,2 0,8 98,3 24,7-168, ,5-98,5-293,5