Operation optimization of existing district heating systems

Operation optimization of existing district heating systems

Applied Thermal Engineering 78 (2015) 278e288 Contents lists available at ScienceDirect Applied Thermal Engineering journal homepage: www.elsevier.c...

1MB Sizes 1 Downloads 113 Views

Applied Thermal Engineering 78 (2015) 278e288

Contents lists available at ScienceDirect

Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng

Research paper

Operation optimization of existing district heating systems Pengfei Jie a, *, Neng Zhu b, Deying Li c a

Department of Building Science, Tsinghua University, Beijing 100084, China School of Environment Science and Engineering, Tianjin University, Tianjin 300072, China c School of Environment and Energy Engineering, Beijing University of Civil Engineering and Architecture, Beijing 100044, China b

h i g h l i g h t s  The  The  The  The

operating cost of existing district heating systems was studied. optimization model used to minimize the operating cost of existing district heating systems was established. operating progress of an existing indirect district heating system was optimized by using the optimization model. operating strategy and heating parameters both influence the optimal results of the existing district heating system.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 July 2014 Accepted 29 December 2014 Available online 6 January 2015

Reducing the operating cost of district heating (DH) systems can be regarded as an operation optimization of DH systems. It was proved that the pumping cost and heat loss cost (PHLC) determined the operating cost of an existing DH system. The minimum operating cost can be obtained by minimizing the PHLC. The optimization method was used. The optimization model used to minimize the PHLC of an existing DH system was established. Program was written to solve the optimization model by using MATLAB software. An existing indirect DH (IDH) system in Hohhot, Inner Mongolia, China was optimized by using the optimization model. Four strategies were used respectively. Results show that the minimum PHLC and related heating parameters can be obtained by using the optimization model. When controlling the primary water mass flow rate (PMF) and the secondary water mass flow rate (SMF) simultaneously, the minimum PHLC is lower than that for other strategies. But this operating strategy requires the excellent hydraulic stability of DH systems. The limit of the frequency of pumps should also be taken into account when applying the optimization model to engineering practice. © 2015 Elsevier Ltd. All rights reserved.

Keywords: District heating Operating cost Optimization model Operating strategy Heating parameters

1. Introduction DH systems have been developed widely because it is considered an effective method to improve the energy efficiency of the space heating system in buildings. DH systems are very common in many countries, such as China, Russia, European countries and so on. In Sweden, energy used in the building sector accounts for 40% of the total energy. 55% of the heat demand of Swedish buildings is supplied by DH systems [1]. In towns in the north of China, the area of buildings connected to DH systems increased from 1.1 billion square meters in 1996 to 9.3 billion square meters in 2011 [2,3]. And therefore, the energy consumption caused by DH systems reached nearly 166 million tons of standard coal in 2011 [2,3]. Urban Persson et al. [4] studied the heat demand in 83 cities in France, Germany,

* Corresponding author. Tel.: þ86 1062788513; fax: þ86 1062770544. E-mail address: [email protected] (P. Jie). http://dx.doi.org/10.1016/j.applthermaleng.2014.12.070 1359-4311/© 2015 Elsevier Ltd. All rights reserved.

Netherlands and Belgium. It was estimated that the average heat market share for DH systems within these 83 cities was 21% in 2006. Combined heat and power plant (CHP) and boilers can be used as heat sources of DH systems. DH systems are changing from fossil fuel-based energy systems to renewable energy systems, which means that fossil fuel (such as coal, oil, natural gas, etc.) is replaced by some non-fossil fuel in the process of heat production [5,6]. Solar, wind, geothermal energy, biomass, etc are used to supply heat for consumers of DH systems [7e14]. Moreover, using waste heat generated during industrial operation or electricity production increases energy efficiency and reduces the use of fossil fuels and other energy resources [15e17]. Therefore, DH systems play an important role in improving indoor thermal comfort and saving primary energy consumption. However, there still remain some problems which should be solved to make DH systems more efficient and more competitive. Heat produced in heat sources needs to be distributed to consumers through DH network. The heat medium can be hot water or steam. In this paper, we focus on DH systems using water as the

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

Nomenclature as, bs Ah b c Chd Chs Cl Cp dw dz Gr h H Kh l m n npp ns nsp Nr Qd Qlp Qls Qs

coefficients for calculating the heat transfer from one piece of radiator to rooms area of heat exchanger, m2 distance between pipe centers, m specific heat of water, kJ/kg operating cost of heat distribution, RMB operating cost of heat sources, RMB heat loss cost, RMB pumping cost, RMB outer pipe diameter, m outer insulation diameter, m pump flow rate at rated speed, kg/s distance between pipe centers and ground surface, m corrected distance between pipe centers and ground surface, m heat transfer coefficient of heat exchanger (W/m2  C) pipe length, m total number of pipe sections with same diameter in the secondary heating network total number of pipe sections with same diameter in the primary heating network total number of circulating pumps in the primary side pieces of radiators total number of circulating pumps in the secondary side pump power at rated speed, kW heat demand of consumers, kW heat loss of the primary heating network, kW heat loss of the secondary heating network, kW heat supply, kW

heat medium. Pumps are required to provide power for the circulation of heat medium in this process. Therefore, the overall cost of heat distribution mainly includes the repayment of capital cost (such as pump investment cost, pipe investment cost, valve investment cost, substation investment cost, etc.), pumping cost, heat loss cost, maintenance cost, salaries, etc. When enough heat is supplied to consumers, how to reduce the total cost is a subject on optimization of DH systems. Some researches on this subject were conducted. Jonas Gustafsson [18] proved that it was possible to use the primary supply water temperature (PST) in the heating network of DH systems to control radiator systems while maintaining the comfort, with the purpose of increasing the temperature difference between PST and primary return water temperature (PRT). Increasing the temperature difference between PST and PRT can reduce the pump electrical energy consumption (PEEC) and improve the overall fuel efficiency. Also, in the CHP plant, more electricity can be produced with colder PRT caused by the increased temperature difference between PST and PRT. Similarly, in order to obtain the lowest PRT, P. Lauenburg et al. [19] used field experiments and computer simulations to develop a control algorithm of radiator systems in DH systems. Henrik Lund et al. [5] pointed out that low-temperature DH systems were beneficial to using low-temperature heat sources (e.g. waste) and improving the utilization efficiency of solar, geothermal energy, etc. Also, the heat loss could be reduced by using lowtemperature DH systems. Aibin Yan et al. [20] developed a hydraulic model to simulate the hydraulic performance of a district heating network with distributed variable speed pumps. They found that such a system could save more energy than the DH system with

Rb1, Rb2

Rc Rt tg Dtm tn tpr tps tsr tss Ue Uh

lb

lt t he hf a b

279

heat resistance for insulation materials of two parallel pipes (supply water pipes and return water pipes) buried in the ground, m  C/W additional heat resistance, m  C/W ground heat resistance, m  C/W ground surface temperature,  C logarithmic mean temperature difference,  C indoor temperature,  C primary return water temperature,  C primary supply water temperature,  C secondary return water temperature,  C secondary supply water temperature,  C electricity price, RMB/kWh heat price, RMB/kWh heat conductivity for insulation materials of heating pipes, W/m  C heat conductivity for the ground, W/m  C a period of time, s efficiency of pump electric motor, % efficiency of pump inverter, % convective heat transfer coefficient between environment and ground surface, W/m2  C additional heat loss coefficient caused by accessories, compensators, valves, etc

Subscripts i number of pipe sections with same diameter in the primary heating network j number of pipe sections with same diameter in the secondary heating network k number of circulating pumps of the primary side q number of circulating pumps of the secondary side min minimum value max maximum value conventional central circulating pumps. K.C.B. Steer et al. [21] studied the influences of the control period on the overall operating cost of DH systems. It was found that an appropriate control policy could save much energy. Tatu Laajalehto [22] proved that the pump power, heat loss and return water temperature could be reduced by utilizing a ring heating network and mass flow rate control, and therefore, significant energy efficiency improvements of DH systems could be achieved. Marouf Pirouti et al. [23] found that the annual energy consumption and equivalent annual cost when using variable flow rate and variable supply water temperature operating strategy was lower compared with other operating methods. The energy consumption of heat sources is determined by heat supply, heat source forms, operating efficiency, etc. The heat supply is mainly determined by the heat demand of consumers. Therefore, as for an existing DH system, the energy consumption of heat sources changes little when the heat supply is constant. The energy consumption of heat distribution is mainly determined by PEEC and heat loss. According to the actual test and statistics in 2005e2006 heating season (from 15 November 2005 to 15 March 2006) in Beijing, the PEEC of direct DH systems is 1.3e2.0 kWh per square meter, while the PEEC of IDH systems is 2.6e3.2 kWh per square meter [24]. Also, it can be seen that much energy can be saved by reducing the PEEC. In this paper, we focus on PEEC and heat loss of IDH systems. The schematic diagram of an IDH system is shown in Fig. 1 [25]. In the primary side, hot water passes to DH substation through primary heating network, and then returns to heat sources. In the secondary side, water obtains heat from hot water in the primary side through heat exchangers, and then heat transfers from water to rooms

280

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

Fig. 1. Schematic diagram of an IDH system.

through radiators. The heat is produced at heat sources. In this process, heating parameters should be optimally selected to effectively reduce the cost of heat distribution for existing DH systems.

Rb ¼

1 dz ln 2plb dw

Rt can be calculated as Eq. (5) [26e28].

2. Methodology In the process of heat distribution, the heating parameters need to be controlled to supply enough heat for consumers. These heating parameters mainly include PST, PRT, secondary supply water temperature (SST), secondary return water temperature (SRT), PMF, and SMF. The water flow rate decreases as the temperature difference between supply water temperature and return water temperature increases, resulting in the decrease of PEEC. But the heating parameters should meet the requirements of heat balance of DH systems.

Rt ¼

1 4H ln 2plt dz

As for IDH systems, there are three kinds of heat balance conditions. Firstly, during a period of time, heat supply is equal to heat demand and heat loss. This can be described as follows.

0

Zt   Qd þ Qlp þ Qls dt Qs dt ¼

(1)

(5)

H can be calculated as Eq. (6) [26e28].

H ¼hþ

lt a

(6)

Rc can be calculated as Eq. (7) [26e28].

2.1. Heat balance equations

Zt

(4)

1 Rc ¼ ln 2plt

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2 2H 1þ b

(7)

Secondly, in heat exchangers, heat transfers to the water in the secondary heating network. The heat transfer process in heat exchangers can be described as follows.

0

When heating pipes are buried in the ground, the heat loss of the heating network can be calculated as follows [26e28].

"

Qlp ¼ 103 ð1 þ bÞ

    n X tps  tg ðRb2i þ Rti Þ  tpr  tg Rci

ðRb1i þ Rti ÞðRb2i þ Rti Þ  R2ci #     n X tpr  tg ðRb1i þ Rti Þ  tps  tg Rci þ li ðRb1i þ Rti ÞðRb2i þ Rti Þ  R2ci i¼1

li

    tss  tg Rb2j þ Rtj  tsr  tg Rcj 3 4    lj Qls ¼ 10 ð1 þ bÞ Rb1j þ Rtj Rb2j þ Rtj  R2cj j¼1   3    m tsr  tg Rb1j þ Rtj  tss  tg Rcj X    þ lj 5 Rb1j þ Rtj Rb2j þ Rtj  R2cj j¼1 



Zt 0:001ns as 0

(3) Rb can be calculated as Eq. (4) [26e28].

ðQd þ Qls Þdt

(8)

0

Thirdly, as for consumers, heat transfers from water to rooms through radiators. The heat balance at radiators can be described as follows.

(2) m X

Zt Kh Ah Dtm dt ¼

0

i¼1

2

Zt

tss þ tsr  tn 2

b s

Zt dt ¼

Qd dt

(9)

0

Eqs. (1)e(9) can be used to describe the heat balance of IDH systems. These equations reflect the cumulative effect of heat transfer. DH systems and buildings both have thermal storage properties [29e32]. If the heat supply is smaller than its needed value at some point, the indoor temperature does not fluctuate too much immediately. Therefore, it is only necessary for the cumulative heat supply in a period of time to be equal to the cumulative heat demand.

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

2.2. Optimization model of an existing IDH system

Cp ¼

How to determine heating parameters is an optimization study. The objective function and some related constraint conditions should be determined. The optimization objective of an existing IDH system is to obtain the minimum operating cost by selecting optimal heating parameters. As for an existing IDH system, the operating cost mainly includes that of heat sources and heat distribution, as shown in Fig. 2. The objective function used to minimize the total operating cost of an existing IDH system can be expressed as follows.

min ðChs þ Chd Þ

(10)

As can be seen in Fig. 2, the operating cost of heat sources changes little when the heat supply remains unchanged. So the heating parameters have little impact on the operating cost of heat sources when the heat supply is constant. Except for the operating cost of heat sources, the operating cost of an existing IDH system mainly includes that of repayment of capital cost, pumping cost, heat loss cost, maintenance cost, salaries, etc. The repayment of capital cost, maintenance cost and salaries remain nearly unchanged as the heating parameters vary. However, the PHLC is related to heating parameters when the heat supply is constant. The total operating cost is mainly determined by the PHLC. Therefore, the minimum total operating cost can be obtained by minimizing the PHLC. When the heat supply is constant, the following objective function can be used to minimize the PHLC of an existing IDH system (also to minimize the total operating cost of an existing IDH system).

  min Cp þ Cl

281

"

(11)

According to the relationship between the flow rate and pump power, the pumping cost can be calculated as Eq. (12) [33].

#3

npp X Ue Nrk tk Q  sk  3600hek hfk Grk c tps  tpr k¼1

3 nsp X Ue Nrq tq Qsq þ 3600heq hfq Grq cðtss  tsr Þ q¼1

(12)

According to Eq. (2) and Eq. (3), the heat loss cost of the heating network can be calculated as Eq. (13).

Cl ¼

"     n tps  tg ðRb2i þ Rti Þ  tpr  tg Rci 103 Uh tð1 þ bÞ X li 3600 ðRb1i þ Rti ÞðRb2i þ Rti Þ  R2ci i¼1 #     n X tpr  tg ðRb1i þ Rti Þ  tps  tg Rci þ li ðRb1i þ Rti ÞðRb2i þ Rti Þ  R2ci i¼1   2    m tss  tg Rb2j þ Rtj  tsr  tg Rcj 103 Uh tð1 þ bÞ 4 X    lj þ 3600 R þR R þ R  R2 j¼1 b1j

tj

b2j

  3    m tsr  tg Rb1j þ Rtj  tss  tg Rcj X    þ lj 5 Rb1j þ Rtj Rb2j þ Rtj  R2cj j¼1

tj

cj

(13) In the process of heat distribution, the heating parameters should meet the requirements of heat balance equations. In addition, the value of SRT, SST, PST, and PRT also has a range. Firstly, increasing the supply water temperature can increase the water temperature difference, resulting in the decrease of PEEC. But if the supply water temperature is too high, it will have an adverse impact on heating pipes and insulation materials. Secondly, in order to improve energy efficiency, we need to select lower return water temperature. Thirdly, water temperature should meet the design requirements. Fourthly, SRT should be lower than PRT. PST should be higher than SST. PRT should be lower than PST. SST

Fig. 2. Illustration of total operation cost of IDH systems.

282

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

should be higher than SRT. SRT should be higher than indoor temperature. Therefore, except for the heat balance constraints (Eqs. (1)e(9)), other constraint conditions used in the optimization model of an existing IDH system are shown as follows.

tsr < tpr

(18)

tss < tps

(19)

tps;min  tps  tps;max

(14)

tpr < tps

(20)

tpr;min  tpr  tpr;max

(15)

tsr < tss

(21)

tn < tsr

(22)

tss;min  tss  tss;max

(16)

tsr;min  tsr  tsr;max

(17)

The above objective function and constraint conditions constitute the optimization model of an existing IDH system. It can be expressed as follows.

8   > min Cp þ Cl > > > > > > s:t: Qs ¼ Qd þ Qlp þ Qls > > > > > > > Kh Ah Dtm ¼ Qd þ Qls > > > >  b s > > > > 0:001n a tss þ tsr  t > ¼ Qd s s n > > 2 > > > " # >         > n n > X X > tps  tg ðRb2i þ Rti Þ  tpr  tg Rci tpr  tg ðRb1i þ Rti Þ  tps  tg Rci > 3 > Qlp ¼ 10 ð1 þ bÞ li þ li > > > ðRb1i þ Rti ÞðRb2i þ Rti Þ  R2ci ðRb1i þ Rti ÞðRb2i þ Rti Þ  R2ci > i¼1 i¼1 > > >     > 2 3       > > > m tss  tg m tsr  tg Rb2j þ Rtj  tsr  tg Rcj Rb1j þ Rtj  tss  tg Rcj X X > > > Q ¼ 103 ð1 þ bÞ4       > lj þ lj 5 ls > > > Rb1j þ Rtj Rb2j þ Rtj  R2cj Rb1j þ Rtj Rb2j þ Rtj  R2cj > j¼1 j¼1 > > > > > > 1 dz > > > ln Rb ¼ > > 2pl d > w b > > > > > 1 4H > > Rt ¼ ln > > > 2plt dz > > > > > > > H ¼ h þ lt > > > a > > sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi > >  2 > > > 1 2H > < Rc ¼ ln 1 þ 2plt b > > #3 " >

3 n nsp > pp X X > Ue Nrq tq Qsq Ue Nrk tk Qsk > > Cp ¼   > þ > > 3600hek hfk Grk c tps  tpr 3600heq hfq Grq cðtss  tsr Þ > > q¼1 k¼1 > > > " #         > > n n > X tps  tg ðRb2i þ Rti Þ  tpr  tg Rci tpr  tg ðRb1i þ Rti Þ  tps  tg Rci 103 Uh tð1 þ bÞ X > > > C ¼ l þ l > l i i > 3600 > ðRb1i þ Rti ÞðRb2i þ Rti Þ  R2ci ðRb1i þ Rti ÞðRb2i þ Rti Þ  R2ci > i¼1 i¼1 > > >     2 3 >       > > m tss  tg m tsr  tg Rb2j þ Rtj  tsr  tg Rcj Rb1j þ Rtj  tss  tg Rcj > X > 103 Uh tð1 þ bÞ 4 X > >       lj þ lj 5 þ > > 3600 > > Rb1j þ Rtj Rb2j þ Rtj  R2cj Rb1j þ Rtj Rb2j þ Rtj  R2cj j¼1 j¼1 > > > > > > tps;min  tps  tps;max > > > > > > > > tpr;min  tpr  tpr;max > > > > > tss;min  tss  tss;max > > > > > > tsr;min  tsr  tsr;max > > > > > > tsr < tpr > > > > > > tss < tps > > > > > > > tpr < tps > > > > > > > tsr < tss > > > :t
sr

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

283

Fig. 3. Program flow chart used to solve the optimization model.

2.3. Solution of the optimization model The above study is an optimization of nonlinear programming problem with constraint conditions. MATLAB software can be used to solve this kind of problem. We can use fmincon function in program [34,35]. The program flow chart used to solve the optimization model is shown in Fig. 3. 3. Results An existing IDH system in Hohhot, Inner Mongolia, China was used as a study case. The operation of this IDH system was optimized. The design outdoor temperature was 17  C, and this temperature was considered as the minimum outdoor temperature in heating season which was from October 20 to April 4. The design indoor temperature was 18  C. The heating area was about

1.78 million square meters. The buildings were all residential buildings. Space heat demand was the only heat demand of this IDH system. The heat demand duration curve is shown in Fig. 4. As shown in Fig. 4, the maximum heat demand of this IDH system was 71.08 MW, and the average heat demand was 47.26 MW. Radiators were installed at rooms so that enough heat transfers from water to rooms through radiators. The design PMF was 1018.64 t/h, and the design SMF was 2444.74 t/h. There was 1 heat source and 14 substations in this IDH system. The diagram of this IDH system is shown in Fig. 5. For the sake of simplicity, the heat source, primary heating network and substations are mentioned in Fig. 5, while the secondary heating network and consumers are not described. The number on pipelines represents the length of heating pipes (m). The red (in the web version) pipelines denote the loop with maximum pressure loss. Fig. 6 describes the pressure of this loop when the PMF keeps the

Fig. 4. Heat demand duration curve.

284

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

Fig. 5. Diagram of an existing IDH system in Hohhot, Inner Mongolia, China.

design value. According to the design requirements, the maximum value of the PST, PRT, SST, and SRT is 130  C, 70  C, 85  C, and 60  C respectively. 3.1. Operating strategy of IDH systems To maintain a stable indoor temperature, enough heat should transfer from water to rooms through radiators. The heat is mainly used for two purposes. Firstly, the heat compensates for the heat

transmission through building envelopes. Secondly, the outdoor air infiltrates into rooms through gaps of doors, windows, etc. The outdoor air should also be heated from outdoor temperature to indoor temperature. So the outdoor temperature is the most important parameter that influences the heat supply. The heat supply must vary as the outdoor temperature varies. Water temperature and water mass flow rate can be controlled to change the heat supply. Four operating strategies based on the variations of these heating parameters are listed in Table 1. In Table 1, constant

Fig. 6. Pressure diagram of the loop with maximum pressure loss when PMF keeps the design value.

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288 Table 1 Operation strategies of IDH systems. Strategy

Primary side

Secondary side

1 2 3 4

Variable flow rate Constant flow rate Constant flow rate Variable flow rate

Constant flow rate Variable flow rate Constant flow rate Variable flow rate

flow rate means that the water mass flow rate is equal to the design parameter, while the water temperature varies in the process of the operation of DH systems; variable flow rate means that the water mass flow rate and water temperature both vary in the process of the operation of DH systems. As different operating strategies may lead to different optimization results, it is necessary to analyze each operating strategy respectively. 3.2. Optimal heating parameters The optimization model was used to optimize the operation of the selected IDH system. Fig. 7 shows the optimal heating parameters for Strategy 1 at different outdoor temperature. As can be seen in Fig. 7, the SMF keeps constant, while the PMF decreases as the outdoor temperature increases. The PRT, SST, and SRT all decrease with the increase of the outdoor temperature, while the PST is different from other heating parameters. The PST keeps constant (130  C) as the outdoor temperature increases from 17  C

285

to 0.8  C. But the PST begins to decrease when the outdoor temperature is above 0.8  C. This is mainly because the heat demand and heat exchange requirements both influence the PST. When the outdoor temperature increases from 17  C to 0.8  C, the heat demand decreases. In this process, the PST can keep the maximum value (130  C) so that the PHLC can be effectively reduced. But when the outdoor temperature is above 0.8  C, the heat demand decreases further. If the PST still keeps the maximum value (130  C), the PHLC may be greater than the minimum PHLC or there may be no correct heating parameters. The variation of the PMF also reflects the similar characteristics. Fig. 8 describes the pressure of the loop with maximum pressure loss when the outdoor temperature is 5  C. Compared with Fig. 6, it can be seen that the pressure decreases when the required PMF decreases. Therefore, the required pump power decreases as the speed decreases. Fig. 9 describes the optimal heating parameters for Strategy 2 at different outdoor temperature. It can be observed that the PMF keeps constant as the outdoor temperature increases. But the SMF decreases as the outdoor temperature increases from 17  C to 2.8  C. When the outdoor temperature is above 2.8  C, the SMF keeps almost unchanged. The PST and PRT both decrease with the increase of the outdoor temperature, while the SST and SRT are different. The SST remains unchanged (85  C) as the outdoor temperature increases from 17  C to 2.8  C. But the SST begins to decrease when the outdoor temperature is above 2.8  C. When the outdoor temperature increases from 17  C to 2.8  C, the heat demand decreases. The SST can keep the maximum value (85  C) so

Fig. 7. Optimal heating parameters for Strategy 1 at different outdoor temperature.

Fig. 8. Pressure diagram of the loop with maximum pressure loss for Strategy 1 (outdoor temperature is 5  C).

286

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

Fig. 9. Optimal heating parameters for Strategy 2 at different outdoor temperature.

that the PHLC can be effectively reduced in this process. But when the outdoor temperature is above 2.8  C, the heat demand decreases further. If the SST still keeps the maximum value (85  C), the PHLC may be greater than the minimum PHLC or there may be no correct heating parameters. The variation of the SRT and SMF also reflect the similar characteristics. Fig. 10 describes the optimal heating parameters for Strategy 3 at different outdoor temperature. The PMF and SMF both remain unchanged as the outdoor temperature increases. The PST, PRT, SST, and SRT all decrease with the increase of the outdoor temperature. Fig. 11 describes the optimal heating parameters for Strategy 4 at different outdoor temperature. Fig. 12 describes the pressure of the loop with maximum pressure loss when the outdoor temperature is 5  C. Compared with Fig. 6, it can be seen that the pressure decreases when the required PMF decreases. Therefore, the pump power decreases as the speed decreases. As can be seen in Fig. 11, the PMF, SMF, and PRT all decrease as the outdoor temperature increases. The PST keeps constant (130  C) as the outdoor temperature increases from 17  C to 2.8  C. When the outdoor temperature is above 2.8  C, the PST begins to decreases. The SST remains unchanged as the outdoor temperature increases from 17  C to 1  C. When the outdoor temperature is above 1  C, the SST begins to decrease. The SRT decreases from 60  C to 20  C as the outdoor temperature increases from 17  C to 1  C. When the outdoor temperature is

above 1  C, the SRT remains unchanged. When the outdoor temperature increases from 17  C to 2.8  C, the heat demand decreases. The PST and SST can both keep the maximum value (130  C and 85  C) so that the PHLC can be effectively reduced in this process. But when the outdoor temperature is above 2.8  C, the heat demand decreases further. If the PST still keeps the maximum value (130  C), the PHLC may be greater than the minimum PHLC or there may be no correct heating parameters. The variation of other parameters also reflects the similar characteristics. 3.3. Minimum PHLC The minimum PHLC for the four strategies is shown in Table 2. The following can be observed: ➢ The minimum PHLC for Strategy 4 is much lower than that for other operating strategies. It means that the PHLC can be effectively reduced by controlling the PMF and SMF simultaneously. However, excellent hydraulic stability of DH systems is required when using Strategy 4. ➢ The minimum PHLC for Strategy 3 is higher than that for other operating strategies. It can be seen that operating strategies with constant mass flow rate of both the primary side and

Fig. 10. Optimal heating parameters for Strategy 3 at different outdoor temperature.

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

287

Fig. 11. Optimal heating parameters for Strategy 4 at different outdoor temperature.

secondary side lead to the highest PHLC. But the DH systems can be operated more steadily when using Strategy 3. This is also beneficial to improving the hydraulic stability of DH systems. ➢ The PHLC for Strategy 2 is lower than that for Strategy 1. Compared with the PHLC when controlling the PMF, more PHLC can be reduced when controlling the SMF. However, the secondary heating network is more complicated than the primary heating network, which increases the difficulty of controlling. According to the heat demand duration curve (Fig. 4) and the optimization model, the minimum annual PHLC for the four strategies can be calculated, as shown in Table 3. Even though the minimum annual heat loss cost for Strategy 4 is not the lowest among that for other strategies, the minimum annual PHLC for Strategy 4 is lower than that for other strategies. As can be seen in Table 3, the minimum annual pumping cost of each operating strategy is different. But the minimum annual heat loss cost for the four strategies varies insignificantly. The operating strategies have little impact on the annual heat loss cost of DH systems. However, the minimum annual pumping cost accounts for a large portion of the minimum annual PHLC. Except for the optimal heating parameters, the minimum annual pumping cost can be effectively reduced by choosing the optimal operating strategy. Therefore,

operating strategies play an important role in reducing the minimum annual PHLC. 4. Conclusion The optimization model was established to minimize the operating cost of an existing IDH system in Hohhot, Inner Mongolia, China. As for an existing DH system, the minimum operating cost can be obtained by minimizing the PHLC. The results show that the minimum PHLC and related optimum heating parameters for the four operating strategies can be obtained by using the optimization model. The optimal results are different. Operating strategies and heating parameters both influence the minimum PHLC. As for the same operating strategy, the minimum PHLC decreases as the outdoor temperature increases. The minimum PHLC for the operating strategies with constant mass flow rate of both the primary side and the secondary side is higher than that for other strategies. But the hydraulic stability of DH systems can be improved by using this strategy. As for Strategy 4, the PMF and SMF both change by using the optimization model, the minimum PHLC is the lowest among that for the four operating strategies. So it can be seen that about 946.44  104 RMB can be saved every heating season when using the optimization model (See Table 3), and this value nearly reflects the best optimization effect. However, the operating

Fig. 12. Pressure diagram of the loop with maximum pressure loss for Strategy 4 (outdoor temperature is 5  C).

288

P. Jie et al. / Applied Thermal Engineering 78 (2015) 278e288

Table 2 Optimal PHLC for different operating strategies (RMB). Outdoor temperature ( C)

Strategy 1

Strategy 2

Strategy 3

Strategy 4

17 15 13 11 9 7 5 3 1 1 3 5

2500.3 2477.9 2462.4 2450.6 2441.5 2434.4 2428.7 2424.1 2420.2 2416.5 2413.1 2409.3

2500.3 1217.2 710.2 449.4 308.1 228.0 181.0 167.3 167.5 168.7 171.6 177.7

2500.3 2498.4 2496.0 2493.5 2491.0 2488.4 2485.8 2483.1 2480.4 2477.7 2474.8 2471.9

2500.3 1192.3 670.9 400.7 252.9 168.6 118.9 88.7 71.0 66.5 61.9 57.2

Table 3 Minimum annual PHLC for different operating strategies (104 RMB). Strategy

1

2

3

4

Annual pumping cost Annual heat loss cost Annual PHLC

952.12 21.64 973.76

55.38 19.41 74.79

976.73 19.75 996.48

29.27 20.77 50.04

strategy with variable PMF and SMF requires the excellent hydraulic stability of DH systems. Compared with the PHLC when only the PMF is controlled, more PHLC can be reduced by controlling the SMF. According to the minimum annual PHLC for the four strategies, the minimum annual pumping cost for each operating strategy is different, while the minimum annual heat loss cost for the four strategies varies insignificantly. It should also be noted that the minimum mass flow of DH systems is related to the limit of the frequency of pumps. So the limit of the frequency of pumps is one of the most important factors which should be taken into account when applying the optimization model to engineering practice. Then the minimum PHLC and optimal heating parameters can be obtained. Acknowledgements The authors would like to thank Dr. Craig Larson for his useful correction of our English language use. The work presented in this paper is supported by “Building Energy Research Center of Tsinghua University” and “Beijing Key Lab of Heating, Gas Supplying, Ventilation and Air Conditioning”, which is hereby greatly acknowledged. We would like to acknowledge the reviewers for their valuable advice on this paper. We also thank the authors of the references in this paper. References [1] Magnus Åberg, Investigating the impact of heat demand reductions on Swedish district heating production using a set of typical system models, Appl. Energy 118 (2014) 246e257. [2] Building Energy Research Center of Tsinghua University, 2012 Annual Report on China Building Energy Efficiency, China Architecture & Building Press, 2012. [3] Building Energy Research Center of Tsinghua University, 2013 Annual Report on China Building Energy Efficiency, China Architecture & Building Press, Beijing, 2013. [4] Urban Persson, Sven Werner, Heat distribution and the future competitiveness of district heating, Appl. Energy 88 (3) (2011) 568e576. [5] Henrik Lund, Sven Werner, Robin Wiltshire, Svend Svendsen, Jan Eric Thorsen, Frede Hvelplund, Brian Vad Mathiesen, 4th generation district heating (4GDH) integrating smart thermal grids into future sustainable energy systems, Energy 68 (2014) 103e111.

[6] Behnaz Rezaie, Marc A. Rosen, District heating and cooling: review of technology and potential enhancements, Appl. Energy 93 (2012) 2e10. [7] Gerhard Faninger, Combined solar-biomass district heating in Australia, Sol. Energy 69 (6) (2000) 425e435. [8] D. Bauer, R. Marx, J. Nußbicker-Lux, F. Ochs, W. Heidemann, H. MüllerSteinhagen, German central solar heating plants with seasonal heat storage, Sol. Energy 84 (4) (2010) 612e613. ki, Extended ordinary differential equation models [9] R. Kicsiny, J. Nagy, Cs. Szalo for solar heating systems with pipes, Appl. Energy 129 (2014) 166e176. [10] X.S. Jiang, Z.X. Jing, Y.Z. Li, Q.H. Wu, W.H. Tang, Modelling and operation optimization of an integrated energy based direct district water-heating system, Energy 64 (2014) 375e388. [11] Poul Alberg Østergaard, Henrik Lund, A renewable energy system in Frederikshavn using low-temperature geothermal energy for district heating, Appl. Energy 88 (2) (2011) 479e487. [12] Leyla Ozgener, Onder Ozgener, Monitoring of energy exergy efficiencies and exergoeconomic parameters of geothermal district heating systems (GDHSs), Appl. Energy 86 (9) (2009) 1704e1711. [13] Ali Keçebas¸, Arif Hepbasl, Conventional and advanced exergoeconomic analyses of geothermal district heating systems, Energy Build. 69 (2014) 434e441. €derstro € m, Biomass gasification in district [14] Elisabeth Wetterlund, Mats So heating systems e the effect of economic energy policies, Appl. Energy 87 (2010) 2914e2922. [15] Augustine N. Ajah, Anish C. Patil, Paulien M. Herder, Johan Grievink, Integrated conceptual design of a robust and reliable waste-heat district heating system, Appl. Therm. Eng. 27(7) 1158e1164. [16] Herena Torío, Dietrich Schmidt, Development of system concepts for improving the performance of a waste heat district heating network with exergy analysis, Energy Build. 42 (10) (2010) 1601e1609. [17] Matteo Morandin, Roman Hackl, Simon Harvey, Economic feasibility of district heating delivery from industrial excess heat: a case study of a Swedish petrochemical cluster, Energy 65 (2014) 209e220. [18] Jonas Gustafsson, Jerker Delsing, Jan Van Deventer, Experimental evaluation of radiator control based on primary supply temperature for district heating substations, Appl. Energy 88 (2011) 4945e4951. [19] P. Lauenburg, J. Wollerstrand, Adaptive control of radiator systems for a lowest possible district heating return temperature, Energy Build. 72 (2014) 132e140. € Jun Huang, [20] Aibin Yan, Jun Zhao, Qingsong An, Yulong Zhao, Hailong Li, Yrjo Hydraulic performance of a new district heating systems with distributed variable speed pumps, Appl. Energy 112 (2013) 876e885. [21] K.C.B. Steer, A. Wirth, S.K. Halgamuge, Control period selection for improved operating performance in district heating networks, Energy Build. 43 (2011) 605e613. €kil€ [22] Tatu Laajalehto, Maunu Kuosa, Tapio Ma a, Markku Lampinen, Risto Lahdelma, Energy efficiency improvements utilising mass flow control and a ring topology in a district heating network, Appl. Therm. Eng. 69 (2014) 86e95. [23] Marouf Pirouti, Audrius Bagdanavicius, Janaka Ekanayake, Jianzhong Wu, Nick Jenkins, Energy consumption and economic analyses of a district heating network, Energy 57 (2013) 149e159. [24] Lanbin Liu, Lin Fu, Yi Jiang, Electrical power consumption analysis of circulating pumps in central heating systems for residential quarters, HV&AC 38 (2008) 123e126. [25] Pengfei Jie, Zhe Tian, Shanshan Yuan, Neng Zhu, Modeling the Dynamic Characteristics of a District Heating Network, vol. 39, 2012, pp. 126e134. [26] Ping He, Gang Sun, Heating Engineering, China Architecture & Building Press, Beijing, China, 1993. [27] Svend Frederiksen, Sven Werner, District Heating and Cooling, Studentlitteratur, Lund, Sweden, 2013. [28] Deying Li, Heating Engineering, China Architecture & Building Press, Beijing, China, 2004.  Aste, Adriana Angelotti, Michela Buzzetti, The influence of the [29] Niccolo external walls thermal inertia on the energy performance of well insulated buildings, Energy Build. 41 (11) (2009) 1181e1187. [30] Koray Ulgen, Experimental and theoretical investigation of effects of wall's thermophysical properties on time lag and decrement factor, Energy Build. 34 (3) (2002) 273e278. [31] Camila Barreneche, Alvaro de Gracia, Susana Serrano, M. Elena Navarro, Ana s Ferna ndez, Manuel Carmona, Juan María Borreguero, A. Ine Francisco Rodriguez, Luisa F. Cabeza, Comparison of three different devices available in Spain to test thermal properties of building materials including phase change materials, Appl. Energy 109 (2013) 421e427. [32] Anna Laura Pisello, Franco Cotana, Andrea Nicolini, Cinzia Buratti, Effect of dynamic characteristics of building envelope on thermal-energy performance in winter conditions: in field experiment, Energy Build. 80 (2014) 218e230. [33] Xiangzhao Fu, Yimin Xiao, Pipe Network for Fluid Distribution, China Architecture & Building Press, Beijing, China, 2010. [34] Sicong Yuan, Jinhua Zhang, Jianxiao Zheng, MATLAB Language and Applied Technology, National Defense Industry Press, Beijing, China, 2011. [35] Jinming Su, Shenyong Ruan, MATLAB Practical Tutorial, Beijing, China, 2008.