A new graphical method to target carbon ... - Wiley Online Library

0 downloads 9 Views 1MB Size Report
we proposed a graphical method that features simultaneous correlation ... CO2 emission reduction, energy saving, fuel switching, furnace, graphical method,.

Received: 10 July 2017

Revised: 6 October 2017

Accepted: 19 October 2017

DOI: 10.1002/er.3946

RESEARCH ARTICLE

A new graphical method to target carbon dioxide emission reductions by simultaneously aligning fuel switching, energy saving, investment cost, carbon credit, and payback time Ahmed Mahmoud

| Jaka Sunarso

Research Centre for Sustainable Technologies, Faculty of Engineering, Computing and Science, Swinburne University of Technology, Jalan Simpang Tiga, 93350 Kuching, Sarawak, Malaysia Correspondence Ahmed Mahmoud, Faculty of Engineering, Computing and Science, Swinburne University of Technology, Jalan Simpang Tiga, 93350 Kuching, Sarawak, Malaysia. Email: [email protected]; [email protected]

Summary Lowering CO2 emissions has become one of the key drivers behind the process intensification and modification in current chemical process industries. Here, we proposed a graphical method that features simultaneous correlation between CO2 emission reduction, fuel switching, energy saving, investment cost, carbon credit, and payback time. Such CO2 emission reduction can be obtained by fuel switching and/or retrofitting of the heat exchanger network. We illustrate the applications of this graphical method to the crude oil preheating train that uses furnace and the palm oil refinery that uses steam boiler. In crude oil preheat train case, for example, 55% emission reduction target can be achieved at an approximately 1.15 year of payback time from the alignment of fuel switching and energy saving. Further reduction in payback time from 1.15 to 0.91 years can be obtained by adding carbon credit contribution scheme into such alignment. This illustrates the flexibility of our graphical method to provide simple and convenient way for evaluating the technical and economic variable relationship for decision‐making. KEYWORDS carbon credit, CO2 emission reduction, energy saving, fuel switching, furnace, graphical method, steam boiler

Nomenclature: a, b, c = cost law coefficients; Cp = specific heat capacity (J kg−1 K−1); C% = mass percentage of carbon in fuel; HTC = heat transfer coefficient (W m−2 K−1); HEN = heat exchanger network; hPROC = enthalpy of steam to process (kJ kg−1); hSUP = enthalpy of superheated steam (kJ kg −1 ); Qfuel = heat duty from fuel (kW); QProc = process duty (kW); LB = latent heat of steam under boiler condition (kJ kg−1); LProc = latent heat of steam delivered to the process (kJ kg−1); NHV = net heating value (kJ kg−1); SB = enthalpy of superheated steam under boiler condition (kJ kg−1); T0 = ambient temperature (°C); TB = condensing temperature of steam under boiler condition (°C); Tevap = feed water temperature as it enters the evaporator section (°C); Tapp = approach temperature which is the difference between condensing temperature of steam under boiler condition, TB, and feed water temperature as it enters the evaporator section, Tevap, (°C); T TFT B = theoretical flame temperature in the boiler (°C); T TFT F = theoretical flame temperature in the furnace (°C); TSTACK = stack temperature (°C); Ts = supply temperature (°C); Tt = target temperature (°C); xDesup= flow rate of desuperheated boiler feed water per kg desuperheated steam; ΔTmin = minimum temperature driving force on composite curve (°C); C% = mole percentage of carbon in fuel; α = the ratio of the molar mass of the oxidized form (CO2) to the non‐oxidized form (C) of the pollutant; ηFurn = furnace efficiency (dimensionless); ΔA = additional surface area requirement; ΔE = energy saving; ΔN = number of additional shells

Int J Energy Res. 2018;42:1551–1562.

wileyonlinelibrary.com/journal/er

Copyright © 2017 John Wiley & Sons, Ltd.

1551

1552

1 | INTRODUCTION Fossil fuel combustion for process heat supply and power generation is presently the largest contributor to the global greenhouse gas emissions. Improvement in the efficiency of the energy systems and switching to the fuel with lower carbon to hydrogen ratio together with carbon dioxide (CO2) capture and storage, and shift to renewable energy resources (such as geothermal, solar, and wind energy) can lower fossil‐fuel based emissions from the industry.1-3 In this context, the European Union Emissions Trading System which was established in 2005 becomes the driver to reduce CO2 emissions from industries through its economic incentives. Using this system, companies can purchase and/or gain emissions allowance, trade this allowance with each other, and purchase finite amount of international credits from emissions‐saving projects all over the world.4 Within the context of minimizing industries' emissions, Smith and Delaby developed models that relate the minimum energy consumption to the (flue gas) emissions from the utilities systems (ie, boiler, furnace, and turbine) for a particular process with certain fixed process conditions.5 They also proposed a methodology to minimize the emissions by the fuel switching, the utilities system design change, the process modification, and the flue gas post‐chemical treatment.6 Varympopiotis et al recently devised a time‐dependent computational model to estimate the potential advantages of the fuel switching application in power plant. Their model predictions highlight significant potential economic benefits for most scenarios that apply fuel switching relative to single‐fuel electricity generation scenario.7 In addition to the fuel switching, the retrofitting of the heat exchanger network (HEN) can be performed to achieve energy saving and emission reduction in chemical process industries. Tjoe formulated a HEN retrofit method based on the capital investment‐energy trade‐off relationship that utilizes pinch analysis.8 They introduced constant area efficiency (α), ie, the ratio of the ideal heat recovery area at the existing energy consumption to the existing heat recovery area of the HEN. Their constant area efficiency curve can be used to predict the additional surface area requirement (ΔA) for any particular energy saving (ΔE) ahead of the actual design process. Following this, the trade‐off relationship curve between the energy saving and the capital investment cost can then be drawn. Tjoe method, which assumes that all retrofit designs maintain the specified area efficiency, has been widely used to target energy saving and studied by several different groups such as Polley et al, Al‐Riyami et al, and Panjeshahi and Tahouni.9-11 Mahmoud et al developed a graphic‐based methodology to reduce the CO2 emissions by combining the fuel

MAHMOUD AND SUNARSO

switching and energy saving via the retrofitting of the HEN.1 Using their method, for any particular emission reduction target, the fuel switching step is firstly applied. After obtaining the CO2 emissions limit, further reduction in emissions can then be achieved by increasing energy saving by retrofitting the HEN at a particular minimum temperature driving force (ΔTmin). Subsequently, Gharaie et al proposed a mathematical model to address CO2 emissions from large‐scale process industries sites.12 Their model takes into account the interactions between process units, associated HENs, and the sites' utilities systems. Their CO2 emission reductions are achieved using the retrofitting of HEN, the optimization of the utility system operation, and the fuel switching with the optimization objective of determining the most suitable CO2‐mitigation options for a particular emission reduction target and an available capital investment by accounting for the carbon trading issue. Their model application into the industry case study demonstrates the significant effect of carbon trading for the optimized configuration, which underlines the economic attractiveness of such carbon trading scheme in real industry scenario. Al‐Mohannadi and Linke developed a new method for the systematic design of low cost carbon integration networks for industrial parks.13 Their method takes into account CO2 management across various sources, utilization and storage options, separation, compression, and transmission options and can assist decision makers to explore carbon footprint reduction options across the industrial park and its vicinity. Hassiba et al later used their carbon dioxide integration method to explore carbon management options across an industrial park together with energy integration approach to minimize net energy demand and fuel consumption of the industrial park.14 They also explored the synergies available from utilizing excess process heat to provide low‐cost, emissions free heat and power sources for energy intensive carbon capture and compression costs. Griffin et al evaluated the opportunities and challenges to reduce industrial energy demand and carbon dioxide emissions in the Chemicals sector by 2050 based on the United Kingdom scenario.15 To achieve significant reduction in carbon emission up to 2050, they suggest the adoption of several key technologies such as carbon capture and storage, energy efficiency techniques, and bioenergy alongside the decarbonization of the electricity supply. Although the mathematical method is considered superior to the graphical method to determine the optimum solution, the latter method provides more advantages in terms of the simple usage and interpretation.12 The objective of this work is to report a new single framework presentation that can align several different

MAHMOUD AND SUNARSO

1553

factors such as fuel switching, energy saving, investment cost, carbon credit, and payback time simultaneously to target CO2 emission reduction in chemical process plants.

follows:

2 | MATHEM ATIC AL M O DEL FOR CO 2 EMISSIONS CALCULATION

CO2 emissions from the furnace can be calculated using Equations 2, 3, and 4.

Carbon dioxide (CO2) is generated mainly from furnaces and boilers that provide heat and steam to the process by fuel combustion. Therefore, furnaces and boilers are the key components for energy saving and CO2 emission reduction. Fuel is combusted by oxygen (O2) in the air, producing CO2 according to the following reaction:  m m Cn H m þ n þ O2 →nCO2 þ H 2 O 4 2

(1)

where n and m represent the mole amount of carbon (C) and hydrogen (H), respectively, that are present in the fuel. Air is assumed to be supplied in excess amount to combustion chambers to ensure complete combustion so that no carbon monoxide is formed. CO2 emissions rate in kg s−1 can then be related to the heat duty of the fuel (ie, coal, fuel oil and natural gas) combusted in furnace and/or boiler, using the model of Smith and Delaby5: CO2 emissions ¼

Qfuel C% × ×α NHV 100

(2)

Qfuel ¼

QProc ηFurn

(4)

2.2 | CO2 emissions from the steam boilers Combustion of fuel is also required to generate steam in boiler. The theoretical flame temperature in the boiler ( T TFT B ) is actually lower relative to that in the furnace because the heat of combustion is transferred immediately to the steam; nonetheless, 1800°C is still adopted as a typical T TFT B value. Steam can be produced in the boiler either at the temperature required by the process or at a higher temperature and then throttled. In the latter case, the steam quality can be maintained by adding another boiler water feed after the expansion. This process is defined as desuperheating and is depicted in Figure 2. If the boiler feed water is assumed to be available at 100°C (h = 419 kJ kg−1), then using heat balance, the mass of boiler feed water required to desuperheat the steam is17: x Desup ¼

hSUP −hPROC hSUP −419

(5)

If an approach temperature, Tapp, of 50°C is assumed in the boiler, the fuel required in the boiler to fulfill the process duty is:

2.1 | CO2 emissions from the furnace Combustion of fuel with oxygen in air produces hot flue gas that can be used for process heating. The maximum amount of heat released by combustion is equal to the amount of heat release required to cool the flue gas from the theoretical flame temperature in the furnace (T TFT F ) to the stack temperature (TSTACK). The T TFT F is generally 1800°C.16 Stack temperature is ideally higher than the corrosion limit. Thus, a typical TSTACK of 160°C is chosen. The flue gas temperature profile as a function of enthalpy for the furnace is shown in Figure 1. The furnace efficiency (ηFurn) is defined as the ratio of the heat accessible to be delivered to the process to the amount of the fuel combusted. ηFurn ¼

T TFT F −T STACK T TFT F −T 0

(3)

The amount of the fuel combusted in the furnace (Qfuel) can then be related to the heat duty required by the process (Qproc) and the furnace efficiency (ηFurn), as

FIGURE 1 Temperature versus enthalpy profile for the furnace flue gas (reproduced with permission from Smith and Delaby5) [Colour figure can be viewed at wileyonlinelibrary.com]

1554

MAHMOUD AND SUNARSO

FIGURE 2 Temperature versus enthalpy profile for the steam raising in the boiler and the steam levels generation by throttling (reproduced with permission from Smith and Delaby5) [Colour figure can be viewed at wileyonlinelibrary.com]

Qfuel ¼

   QPROC  T TFT B −T 0 1−x Desup ðLB þ SB Þ LProc T TFT B −T B −50

(6)

CO2 emissions from the boiler can be calculated using Equations 2, 5, and 6.

3 | RETROFITTING OF HEAT EXCHANGER NETWORK In retrofitting the HEN to achieve energy saving, generally installation of additional heat exchanger surface area is required. The widely used method for determining this additional area is the Pinch Design method, which we describe briefly later. The minimum energy requirement and the minimum area requirement are functions of minimum temperature driving force (ΔTmin). Figure 3 shows a typical energy‐ area relationship plot for a given data set of streams. Point B on Figure 3 indicates the optimum trade‐off point for a new HEN configuration, while Point X represents this point for the existing HEN configuration. In retrofitting scenario nonetheless, Figure 3 considers only heat recovery area from process‐to‐process heat exchangers. Upon comparing the current plant performance (Point X) with the optimum grass‐root design (Point B), the optimum grass‐root design displays lower energy and lower area requirements than the existing network. Retrofitting to change the condition from Point X to Point B necessitates the removal of a portion of heat exchange area that has been installed and is available at no extra cost. To this

end, Tjoe mentioned that better retrofitting should make good use of the investment that has already been made.8 Ideal retrofitting should move the process condition from Point X (the existing inefficient condition) to Point A (the most efficient one that can be achieved). In principle, such process modification is possible, but in practice, it is often undesirable. The network structure that represents Point A may be so different with the initial network configuration. This may require different heat exchanger sizes and types. Because any modifications have associated investment cost, it is more reasonable to assume that such modification will move Point X to Point Z. Determining the optimum retrofitting path is not straightforward. The network after retrofitting should at least use heat exchange area as effectively as the existing network. If the current placement is already appropriate, the new area should not be created in a way that reduces the effectiveness of the surface area in the network. Thus, area efficiency (α) is defined as the ratio of the minimum area requirement (target area) to the area that is actually used in the existing network for the existing energy requirement. The demonstration of area efficiency (α) and the addition of retrofit area are explained with more details in the Appendix A.

4 | N E W GR A P H I C A L A LI G N M E N T METHOD Our method that aligns fuel switching, energy saving, investment cost, carbon credit, and payback time to target CO2 emission reduction in chemical processes is demonstrated using two following case studies. Case study 1 utilizes the furnace to provide process heating while Case study 2 utilizes the steam boiler to provide process heating.

FIGURE 3

Energy versus area for a given data set of streams

MAHMOUD AND SUNARSO

1555

Table 1 displays the composition and net heating value (NHV) of the available fuels for these two case studies. For the base case, the existing fuel is assumed to be fuel oil that is present at 25°C. The air for combustion is also supplied at 25°C. Any CO2 emission reduction is credited at a rate of 4.5 £ t−1 of CO2.18

4.1 | Case study 1 The processes for this case study focus on the crude oil preheating train that contains of a furnace, a desalter, and a distillation tower, displayed in Figure 4.11 The crude TABLE 1

Fuel compositions and net heating values

Composition

Oil

Natural Gas

Carbon

87.26

76.0

Hydrogen

10.94

22.8

Oxygen

0.64



Nitrogen

0.28

1.1

Sulfur

0.84

0.1

0.04



Ash −1

NHV (kJ kg )

39 830

Note: Fuel compositions are given in mass %.

51 550

oil is supplied to the distillation tower from the storage at the ambient temperature. This oil is preheated at two different sections by heat exchange with the hot streams coming out from the distillation tower. The first section runs from the storage to the desalter unit. The second section runs from the desalter to the distillation tower. The furnace provides the process heating that heats the crude oil right before entering the distillation tower. Any improvement in the heat recovery within the heat exchange network (HEN) reduces the amount of the external heat required by the process, thus leading to the lower fuel requirement in the furnace and also emission reduction. Figure 5 in turn displays the grid diagram for the existing HEN.11 The hot streams are grouped together in the upper section of the grid and are labeled as H1 to H6. Crude oil feed is the only cold stream, which is labeled as C1. Table 2 lists the temperature, flow, and heat transfer coefficient (HTC) data for these 7 streams and the associated costs (in the note below it), while Table 3 lists the existing heat loads and heat transfer areas for the 6 available heat exchangers. The heat requirement for the configuration shown in Figure 5 is 80 418 kW. This heat duty can be converted into emission flow rate using Equations 2, 3, and 4.

23

TABLE 2 Temperature, flow, and heat transfer coefficient (HTC) data for 7 different streams and the associated costs

FIGURE 4

FIGURE 5

Process flowsheet for the crude oil preheating train

The grid diagram of the existing heat exchanger network for the crude oil preheating train

Stream

Flow, kg s−1

Ts, °C

Tt, °C

HTC, W m−2 K−1

H1

23

180

30

492.2

H2

44

270

40

477.8

H3

13

350

30

439.8

H4

56

380

50

470.7

H5

253

150

100

561.5

H6

148

290

190

432.6

C1

200

20

390

343.0

Note: Heat exchanger capital cost (£) = 8600 + 670 (area)0.83. Hot utility cost (£ kW−1 yr−1) = 70. Cold utility cost (£ kW−1 yr−1) = 7. Maximum area per shell = 580 m2. CP = 2600 J kg−1 K−1 for all streams.

1556

MAHMOUD AND SUNARSO

TABLE 3 The existing heat loads and heat transfer areas for the 6 available heat exchangers Heat Exchanger

Heat Load, kW

Heat Transfer Area, m2

E1

6000

280

E2

23 000

1480

E3

750

280

E4

15 000

800

E5

38 480

2760

E6

22 000

1360

The mass flow rate of a certain emissions product can be calculated based on the fuel type. The reduction in CO2 emissions by fuel switching from oil to natural gas at the current heat consumption of 80 418 kW is 32.7%. Such reduction is attributed to the lower carbon to hydrogen ratio and higher net heating value (NHV) for the natural gas relative to the oil (Table 1). Figures 6 and 7 display

FIGURE 6

CO2 emission reduction by fuel switching to natural gas from oil at different heat duty values based on any projected heat duty value

FIGURE 7

CO2 emission reduction by fuel switching to natural gas from oil at different heat duty values based on the existing heat duty value of 80 418 kW

two different scenarios that picture the CO2 emission reduction as a function of heat duty where the first one is based on any projected heat duty value while the second one is based on the existing heat duty value. In Figure 6, upon switching to the natural gas, a constant CO2 emission reduction (of 32.7%) is maintained for any heat duty value, ie, constant and identical slope for both oil and natural gas cases. This is obtained if the CO2 emission reduction at a particular heat duty is calculated as the difference between CO2 emission flow rate using the oil and CO2 emission flow rate using the natural gas divided by CO2 emissions flow rate using the oil at this same particular heat duty, ie, at heat duty of 8.7·104, CO2%↓ = 32.7% and at heat duty of 5·104, CO2%↓ = 32.7%. In Figure 7, however, the slope for oil case is different with that for natural case, which leads to different CO2 emission reduction at different heat duty value. This is because the difference between CO2 emission flow rate using the oil and CO2 emissions flow rate using the natural gas obtained at any particular heat duty value is divided by CO2 emissions flow rate using oil at the existing heat duty of 80 418 kW. For example, at heat duty of 8.7·104, CO2%↓ = 32.7% and at heat duty of 5.0·104, CO2%↓ = 18.8%. Using the scenario represented in Figure 7, Figure 8 illustrates how fuel switching can be combined with retrofitting of HEN to reduce CO2 emissions further. For a CO2 emission reduction target of 55% (A), fuel switching from oil to natural gas only achieves 32.7% CO2 emission reduction (B), lacking 22.3% extra CO2 emission reduction (C) that can be fulfilled using retrofitting of HEN. The intersection of the lowest dotted horizontal line with the operating line for natural gas provides the new heat duty value required to obtain 22.3% extra CO2 emission reduction. Therefore, the present heat duty of natural gas should be reduced by 33% (D) to reach such reduction. Note that the overall CO2 emission reduction itself is not affected by the sequence of the modification steps. Fuel switching can be carried out before retrofitting of HEN or after and the overall reduction stays the same. As discussed earlier, when we start with fuel switching, 32.7% CO2 emission reduction is achieved. To obtain 55% overall CO2 emission reduction target, extra 22.3% CO2 emission reduction should be achieved via retrofitting HEN and requires reduction in heat duty by 33%. On the other hand, when we start with retrofitting of HEN, by reducing the heat duty by 33%, we obtain 32.7% CO2 emission reduction. We are then left with 22.3% CO2 emission reduction that should be achieved via fuel switching. In both cases, 33% reduction in heat duty should be attained. The notable difference is that in fuel switching before retrofitting HEN case, CO2 emission reduction is 32.7%, while in retrofitting HEN before fuel

MAHMOUD AND SUNARSO

1557

FIGURE 8

Graphical illustration on how to combine fuel switching with retrofitting of heat exchange network to obtain a particular target of CO2 emission reduction for case study 1

switching case, CO2 emission reduction is 22.3%. This is because for the latter case, the difference between CO2 emissions flow rate using the oil and CO2 emissions flow rate using the natural gas obtained at lower heat duty of 53 768 kW after the retrofit of HEN is divided by CO2 emission flow rate using oil at the existing heat duty of 80 418 kW. In the former case, however, the difference between CO2 emission flow rate using the oil and CO2 emission flow rate using the natural gas obtained at the existing heat duty of 80 418 kW is divided by CO2 emission flow rate using oil at the same heat duty of 80 418 kW. Table 3 reveals that the existing HEN contains an overall process‐to‐process heat exchange area of 6960 m2 and consumes an overall hot utility load of 80 418 kW. The ideal heat exchange area required to fulfill the same load value is 5755 m2. Therefore, the existing heat exchange area efficiency is 0.827. Upon knowing the available heat exchange area, the retrofitting curve displayed in Figure 9 can be used (see Appendix A on how to construct such curve), which illustrates how additional area (ΔA) can be added to achieve a particular energy saving (ΔE). For example, based on the existing HEN design, if 33% additional energy saving (ΔE) is to be achieved (over the x‐axis), then additional area (ΔA) of 9991 m2 at

FIGURE 9 heat load

Retrofit curve that relates heat exchange area with

minimum temperature driving force (ΔTmin) of 32.5°C should be provided.

FIGURE 10 Graphical method that allows simultaneous correlation between CO2 emission reduction, fuel switching, energy saving, investment cost, carbon credit, and payback time obtained by combining fuel switching and retrofitting of heat exchanger network for case study 1 [Colour figure can be viewed at wileyonlinelibrary.com]

1558

MAHMOUD AND SUNARSO

FIGURE 11 The grid diagram of the existing heat exchanger network for the palm oil refinery. Note the labels and units for the two subsequent numbers are Q (kW)\A (m2)

Next, we assume that the total investment from fuel switching and retrofitting of HEN can be approximated by the investment of retrofitting of HEN.1 This assumption is justifiable because in several cases, changing the combustion device burner is the only major modification required to accommodate the different fuel and the cost of such change is marginal in comparison to retrofitting cost. Figure 10 illustrates our graphical method to simultaneously measure the CO2 emission reduction obtained from aligning fuel switching, energy saving, investment cost, carbon credit, and payback time. This graphical method provides a convenient way to evaluate the effect of the certain variables for decision‐making purpose that relates chemical engineering technical factors with economic factors. For example, if we set the CO2 emission reduction target to 55%, then the other respective variables, such as energy saving of 33%, the required investment of 2.4·106 £, carbon credit of 5.4·105 £ yr−1, payback time for energy saving of 1.15 years, and payback time for energy saving and carbon credit of 0.91 year can be obtained by simple inspection.

palm oil refinery case. The hot streams are lumped together in the upper section of the grid and are labeled as H1 to H4. The cold streams are lumped together in the bottom section of the grid and are labeled as C1 to C3. The required process heating is provided by medium pressure (MP) and low pressure (LP) steams. Table 4 lists the supply temperature, target temperature, HTC, and cost data for the utilities, ie, cooling water, LP steam, and MP steam. The total heat capacity (CP, which is the product of mass flowrate (m) and specific heat capacity (cp)) for each stream is calculated from the grid diagram of HEN (Figure 11) by dividing the heat load of each stream with the absolute magnitude of the temperature difference between the source and target. The HTC of the 7 palm oil streams is 0.2596 W m−2 K−1.21 This case study has been adopted previously in the work of Mahmoud et al.22 The existing HEN configuration was designed using a minimum temperature driving force (ΔTmin) of 30°C. In our scenario, after a period of operation, the heat exchangers experience fouling associated with the processing of edible oil, resulting in the significant increase in the energy consumption of HEN. Figure 12 displays

4.2 | Case study 2 The focus of this case study is the existing HEN in palm oil refinery as reported in the work of Manan and Yusof.19 The details on the process flowsheet for palm oil refinery can be found in the work of Haslenda et al.20 Figure 11 displays the grid diagram for the existing HEN in our TABLE 4 Target temperatures, supply temperatures, heat transfer coefficients, and costs for 3 different utility streams Type

Tsupply, °C

Ttarget, °C

Hf, kW m−2 K−1

$ kW−1 yr−1

Cooling water

25

30

2.5

18.2

LP steam

143

142

4.5

106.4

MP steam

184

183

4.5

159.6

Note: heat exchanger capital cost = 1054 A

0.65

(A = heat exchange area).

FIGURE 12

Balance grand composite curve of the existing heat exchanger network for the palm oil refinery with ΔTmin of 30°C (before fouling) [Colour figure can be viewed at wileyonlinelibrary. com]

MAHMOUD AND SUNARSO

FIGURE 13 Balance grand composite curve of the existing heat exchanger network for the palm oil refinery with minimum temperature driving force (ΔTmin) of 49.03°C (after fouling) [Colour figure can be viewed at wileyonlinelibrary.com]

the balance grand composite curve (BGCC) of the existing HEN before fouling assuming ΔTmin of 30°C. Figure 13 shows the BGCC after fouling takes place. Upon comparing Figures 12 and 13, it becomes clear that energy consumption after fouling requires external heat source. After fouling, ΔTmin becomes 49.03°C, which is 44% greater than the existing design using ΔTmin of 30°C. As such, additional heating and cooling from using external sources are required (see the dotted upper and bottom lines in Figure 13). Note that fouling phenomenon is generally a sensitive function of temperature. It starts to occur significantly above 200°C.19 Two streams are directly exposed to temperature higher than 200°C, ie, the hot stream H2 and the cold stream C3 (Figure 11). The common solution to solve fouling in heat exchanger is by cleaning. Let us assume that after such cleaning, we can recover the design conditions of ΔTmin of 30°C, energy consumption of 311.24 kW, area efficiency of 0.88, and CO2 emission flow rate of 88 kg hr−1. Figure 14 shows the energy requirement versus the minimum temperature driving force (ΔTmin). The increase in

1559

the energy requirement varies at different ΔTmin ranges where within one range, the change in ΔTmin leads to significant change in energy and within the other ranges, such change leads to lower change in energy. The change in the energy requirement and the associated energy cost is considerable between 23.5°C and 49°C. Between 16°C and 23.5°C, nonetheless, the energy changes marginally with the increase in ΔTmin. Therefore, we pick ΔTmin of 17°C as the target for the retrofit design to maintain lower energy requirement. At such ΔTmin, any temperature change on the process stream translates to small change in energy requirement. To attain the targeted CO2 emission reduction, we start by retrofitting of HEN using ΔTmin of 17°C followed by fuel switching from oil to natural gas. Figure 15 shows that retrofitting of HEN using ΔTmin of 17°C leads to energy saving of 52% (A) and also CO2 emission reduction of 52% (B). The additional fuel switching step provides an extra CO2 emission reduction of 16% (C). The combined retrofitting of HEN and fuel switching thus provides an overall CO2 emission reduction of 68% (D).

FIGURE 15

Graphical illustration on how to combine fuel switching with retrofitting of heat exchange network to obtain a particular target of CO2 emission reduction for case study 2 [Colour figure can be viewed at wileyonlinelibrary.com]

FIGURE 14 Energy requirement of the existing heat exchanger network for the palm oil refinery as a function of minimum temperature driving force (ΔTmin) [Colour figure can be viewed at wileyonlinelibrary.com]

1560

FIGURE 16

Graphical method that allows simultaneous correlation between CO2 emission reduction, fuel switching, energy saving, investment cost, carbon credit, and payback time obtained by combining fuel switching and retrofitting of heat exchanger network for case study 2 [Colour figure can be viewed at wileyonlinelibrary.com]

We have illustrated above single case based on ΔTmin of 17°C although the HEN can actually be retrofitted over a particular range of ΔTmin, ie, between 0°C and 30°C to provide some more degrees of freedom to set different energy saving and CO2 emission reduction targets. Figure 16 then illustrates how CO2 emission reduction from the combined retrofitting of HEN and fuel switching can be acquired simultaneously with fuel switching, energy saving, investment cost, carbon credit, and payback time in a single graph. For example, at ΔTmin of 17°C, the energy saving is 52%, the respective CO2 emission reduction is 68% with the required investment of 4.5·10−4 £, carbon credit of 2.3·103 £ yr−1, payback time for energy saving of 1 year, and payback time for energy saving and carbon credit of 0.96 year.

MAHMOUD AND SUNARSO

FIGURE 17 Graphical method that allows simultaneous correlation between CO2 emission reduction, energy saving, investment cost, carbon credit, and payback time obtained by retrofitting of heat exchanger network only for case study 2 [Colour figure can be viewed at wileyonlinelibrary.com]

If the existing fuel is natural gas and the only possible further modification is to reduce CO2 emissions by retrofitting of HEN, then the energy saving from such retrofitting together with the investment cost, the carbon credit, and the payback time can be obtained simultaneously as illustrated in Figure 17. For energy saving of 52%, the corresponding CO2 emission reduction is 52% with the required investment of 4.5·10−4 £, carbon credit of 1.7·103 £ yr−1, payback time for energy saving of 1 year, and payback time for energy saving and carbon credit of 0.97 year. Compared to retrofitting of HEN by itself, the combined retrofitting of HEN and fuel switching achieves higher CO2 emission reduction and hence, higher carbon credit rate. This is because carbon credit rate increases linearly with the increase in emission reduction.

MAHMOUD AND SUNARSO

5 | CONCLUSION This work proposed a new graphical method for targeting CO2 emission reduction by simultaneously aligning fuel switching, energy saving, investment cost, carbon credit, and payback time in a single graph. We have demonstrated the efficacy of our method, ie, the flexibility to achieve CO2 emission reduction target on preheat train of crude distillation unit and palm oil refinery case studies. For instance, to achieve 55% CO2 emission reduction target in crude oil preheat train, 32.7% of the emission can be achieved via fuel switching while the remaining 22.3% can be achieved via the retrofit of HEN. In the second case, the retrofitting of HEN of palm oil refinery at ΔTmin of 17°C provides 52% savings in energy, which is equivalent to 52% reduction in CO2 emissions. Further CO2 emission reduction up to 68% can additionally be achieved by combining the retrofitting of HEN with fuel switching. Moreover, the inclusion of carbon credit contribution scheme led to 20% and 4% reduction in payback time of the crude oil preheat train case and palm oil refinery case, respectively. What we have demonstrated highlights the capability of the proposed method to provide insights on the relationships of technical and economic variables for decision‐making.

ORCID Ahmed Mahmoud http://orcid.org/0000-0001-8841-8679 Jaka Sunarso http://orcid.org/0000-0002-5234-7431 R EF E RE N C E S 1. Mahmoud A, Shuhaimi M, Abdel‐Samed M. A combined process integration and fuel switching strategy for emissions reduction in chemical process plants. Energy. 2009;34(2):190‐195. https:// doi.org/10.1016/j.energy.2008.11.007 2. Olajire A. CO2 capture and separation technologies for end‐of‐ pipe applications—a review. Energy. 2010; 35: 2610–2628, CO2 capture and separation technologies for end‐of‐pipe applications – A review, 6, DOI: https://doi.org/10.1016/j.energy.2010.02.030. 3. Varbanov P, Perry S, Klemes J, Smith R. Synthesis of industrial utility systems: cost‐effective de‐carbonization. Appl Therm Engineering. 2005;25(7):985‐1001. https://doi.org/10.1016/j. applthermaleng.2004.06.023 4. European commission, The EU Emissions Trading System (EU ETS), http://ec.europa.eu/clima/policies/ets_en (Accessed 26.05. 2017).

1561

7. Varympopiotis G, Tolis A, Rentizelas A. Fuel switching in power‐plants: modelling and impact on the analysis of energy projects. Energ Conver Manage. 2014;77:650‐667. https://doi. org/10.1016/j.enconman.2013.10.032 8. Tjoe T. Retrofit of heat exchanger network. PhD Thesis. UMIST, UK, 1986. 9. Polley G, PanjehShahi M, Jegede F. Pressure drop considerations in the retrofit of heat exchanger networks. Trans IChemE. 1990;68:211‐220. 10. Al‐Riyami BA, Klemes J, Perry S. Heat integration retrofit analysis of a heat exchanger network of a fluid catalytic cracking plant. Appl Therm Eng. 2001;21(13–14):1449‐1487. https://doi. org/10.1016/S1359‐4311(01)00028‐X 11. Panjeshahi M, Tahouni N. Pressure drop optimisation in debottlenecking of heat exchanger networks. Energy. 2008;33(6):942‐951. https://doi.org/10.1016/j.energy.2007.09.013 12. Gharaie M, Zhang N, Jobson M, Smith R, Panjeshahi HM. Simultaneous optimization of CO2 emissions reduction strategies for effective carbon control in the process industries. Chem Eng Res Des. 2013;9(1):1483‐1498. 13. Al‐Mohannadi DM, Linke P. On the systematic carbon integration of industrial parks for climate footprint reduction. J Clean Prod. 2016;112:4053‐4064. https://doi.org/10.1016/j.jclepro.2015.05.094 14. Hassiba RJ, Al‐Mohannadi DM, Linke P. Carbon dioxide and heat integration of industrial parks. J Clean Prod. 2017;155:47‐56. https://doi.org/10.1016/j.jclepro.2016.09.094 15. Griffin PW, Hammond GP, Norman JB. Opportunities for energy demand and carbon emissions reduction in the chemicals sector. Energy Procedia. 2017;105:4347‐4356. https://doi.org/10.1016/j. egypro.2017.03.913 16. Linnhoff B, de Leur J. Appropriate placement of furnaces in the integrated process. In: The University of Manchester, Institute of Science and Technology, UK, 23–23 March 1988: The Institution of Chemical Engineers, Rugby, UK, (1988). p. 258–282. 17. Townsend D, Linnhoff B. Heat and power networks in process design. Part II: design procedure for equipment selection and process matching. AIChE J. 1983;29(5):748‐771. https://doi.org/ 10.1002/aic.690290509 18. European Energy Exchange, 2017. EU emissions allowances, https://www.eex.com/en/market‐data/environmental‐markets/ spot‐market/european‐emission‐allowances#!/2017/05/26 (Accessed 26.05.2017). 19. Manan Z, Yusof F. Retrofit of palm oil refinery to improve energy consumption and reduce fouling costs. Seminar Penyelidikan, FKKSA, UTM. Johor, Malaysia, 1994. 20. Haslenda H, Shuhaimi M, Woo S, Mahmoud A. Operation system optimization. In: Zheng QP, Rebennack S, Pardalos PM, Pereira MV, Iliadis NA, eds. Handbook of CO2 in Power Systems. Heidelberg: Springer; 2012:373‐391.

5. Smith R, Delaby O. Targeting flue gas emissions. Trans IChemE. 1991;69:492‐505.

21. Miller K, Singh R, Franks B. Viscosity and heat transfer coefficient for canola, corn, palm and soybean oil. J Food Proc Preservation. 1994;18(6):461‐472. https://doi.org/10.1111/j.1745‐ 4549.1994.tb00268.x

6. Delaby O, Smith R. Minimization of flue gas emissions. Trans IChemE. 1995;73:21‐32.

22. Mahmoud A, Shuhaimi M, Abdel‐Samed M. Emissions reduction strategy using process integration approach. 15th

1562

MAHMOUD AND SUNARSO

Regional Symposium on Chemical Engineering in Conjunction with 22nd Symposium of Malaysian Chemical Engineers (RSCE‐SOMCHE). Kuala Lumpur, Malaysia, December, 2008. 23. Shenoy V. Heat Exchanger Network Synthesis, Process Optimization by Energy and Resource Analysis. Texas: Gulf Publishing Company; 1995.

How to cite this article: Mahmoud A, Sunarso J. A new graphical method to target carbon dioxide emission reductions by simultaneously aligning fuel switching, energy saving, investment cost, carbon credit, and payback time. Int J Energy Res. 2018;42:1551–1562. https://doi.org/10.1002/er.3946

A P P EN D I X A

A1 | Retrofit and saving‐investment curves If area efficiency (α) is assumed to be constant over the complete energy range, a constant α curve, such as XY curve in Figure A1 is obtained. Such constant α curve

(retrofit curve) enables the engineers to determine the required additional surface area requirement (ΔA) to obtain a particular energy saving (ΔE) ahead of the real design. Upon knowing such value, the investment cost can then be calculated from Equation A1 by assuming that new exchangers have an identical area to the existing network. 

ΔA Investment ¼ ΔN a þ b ΔN

Retrofit target curve

(A1)

Where a, b, and c are the cost law coefficients that depend on the construction materials, the pressure rating, and the heat exchanger type; ΔA is the required additional area to achieve the energy saving target; and ΔN is the required number of additional shells. For any retrofit target curve, the incremental investment in the form of additional area (ΔA) can be correlated to the incremental saving in the form of energy (ΔE) as depicted in Figure A2. By converting both of them into economic factors, the criterion of the project viability such as payback time or investment limit can be determined (Figure A2).

FIGURE A2 FIGURE A1

C !

Saving versus investment curve

Suggest Documents