1、Shinya Nagae, Shiori Takamura, and Yoshitaka Uno are graduate students and Yoshiyuki Shimoda is a professor in the Division of Sustainable Energy and Environmental Engineering, Graduate School of Environment, Osaka University, Osaka, Japan. Kenichiro Watanabe is a professor, Shibaura Institute of Te
2、chnology University, Tokyo, Japan. Yutaka Shoji is a co-researcher in Energy Advance CO., Ltd., Tokyo, Japan. Verification of the Energy Efficiency Advancement in District Heating and Cooling Plant by Renovation Shinya Nagae Yoshiyuki Shimoda, PhD Shiori Takamura Yoshitaka Uno Kenichiro Watanabe Yut
3、aka Shoji ABSTRACT Energy efficiency improvements of renovated district heating and cooling (DHC) plants were evaluated by simulation. In this paper, simulation models for the original and renovated plants were developed using the equipment specifications of the plant before and after renovation. Th
4、e model accuracies were examined by comparing with measured data from plant operations. This comparison resulted in modifications of certain parameters related to the chiller operational control and chiller efficiency. The simulation model quantified the total annual energy efficiency improvement an
5、d the contribution of each piece of equipment replaced. INTRODUCTION In Japan, district heating and cooling systems (DHC) have been in use for approximately 40 years, and about 150 plants have been constructed as high-efficiency heat supply systems in central business districts. DHC plants are class
6、ified into the following three categories on the basis of the energy source: electric heat pump driven systems, which run on electricity; absorption chiller and boiler systems, which use natural gas; and systems that are combinations of these two types. Even within a single category, measured result
7、s show that the energy efficiencies of DHC plants vary widely due to differences in factors such as the heat demand profile, efficiency of the heat source machines, system design, and operation. However, a simulation study proved that DHC plants usually show higher energy efficiency than conventiona
8、l heat source systems in individual buildings because of the concentration effect of heat demand and the grade of operation (Shimoda et al. 2008). In addition, in absorption chiller and boiler DHC systems, the introduction of combined heat and power (CHP) has a unique advantage for energy efficiency
9、 improvement. In recent years, the energy efficiency of chillers, electricity generators, and pumps related to DHC systems have progressed remarkably. This indicates that there is a potential for significant improvements in the total energy efficiency of DHC systems by renovating the plants. For pla
10、nts with an absorption chiller and boiler with CHP, introduction of large-scale, high-efficiency electricity generators, such as gas engines, and high-efficiency turbo refrigerators, which enable the CHP system to operate for longer hours, is expected to significantly increase the energy efficiency
11、of the DHC system (Kubara et al. 2007). In addition, operating improvements also affect DHC plant efficiency (Wang et al. 2007, Ono et al. 2007). A plant with an absorption chiller and boiler with CHP was chosen for our case study. The plant was originally constructed in 1992 and renovated in 2008.
12、In this paper, the simulation models for both the original and the renovated plant were developed using the equipment specifications of the plant before and after renovation. The model accuracy was examined by comparing with measured data. From this comparison, some parameters related to chiller ope
13、ration sequence control and chiller performance were modified. The total annual energy efficiency LV-11-C017 2011 ASHRAE 1392011. American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions, Volume 117, Part 1. For personal use o
14、nly. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAES prior written permission.improvement and the contribution of each piece of equipment replaced were quantified using the model. OUTLINE OF THE DHC PLANT AND RENOVATION WORK Thi
15、s plant is located in central Tokyo, Japan. It has 22.6 MW (77.1 MBtu/h) cooling capacity and 30.1 MW (103 MBtu/h) heating capacity for supplying heat to four buildings, including a hospital, an office, and an apartment. Initially, this plant had two gas-engine CHP systems. They supplied electricity
16、 to the hospital, and waste heat, which consists of low-pressure (0.09 MPa (0.9 bar) and high-pressure (0.78 MPa (7.8 bar) steam, to the DHC plant. Three gas-fired boilers produced high-pressure steam. Cooling heat was produced by two single- and double-effect absorption chillers, which consumed bot
17、h high- and low-pressure steam, and four double-effect absorption chillers, which consumed high-pressure steam. During renovation, the CHP system was replaced with a highly efficient gas engine generator that supplies electricity to the hospital, and waste heat, which consists of high-pressure steam
18、 and hot water (88C (190.4F), to the DHC plant. The single- and double-effect absorption chillers were replaced with a high-efficiency waste-heat utilization absorption chiller, which consumes hot water or high-pressure steam, or both, and a high-efficiency variable speed turbo chiller. Two of the f
19、our double-effect absorption chillers were replaced with brand-new high-efficiency models. Table 1 lists the heat source equipment used before and after renovation. In addition, two cold water pumps and two cooling wa0ter pumps were replaced with variable speed inverter driven pumps. Figure 1 shows
20、the system diagram after renovation. C1C2C5C6C3C4SUPPLYHEADERRETURNHEADERCT1CT2CT5CT6CT3CT4BYPASSCooling demand and flow rateElectricity generation by gas engineGas engine performance modelHot water supplyCity gas demandChiller sequence control modelChiller performance modelChiller ON/OFF and load f
21、actorSteam supplyCooling tower modelHeating (Steam) demandBoiler performance modelWeather data Cooling water temp.Electricity demandInput dataSIMULATION MODEL FOR ORIGINAL PLANT Numerical models were developed for simulating the energy consumption of this plant before and after renovation. Figure 2
22、shows the flowchart of the model. The time step of this model is one hour. The chiller sequence control model determines the chillers that must be operated to meet both the required heat load and flow rate. When the gas engine is operated, at least one chiller that uses waste heat (hot water or low-
23、pressure steam) is in operation. In the 0.01.0Cold heat demandLoad factorofchiller2nd chiller3rd chillerDead band margin1st chillerFigure 3 Dead band marginTable 1. List of heat source equipmentsBEFORE RENOVATIONAFTER RENOVATIONGas engine generator480 kW 2(1.64 106Btu/h 2) = 29.0% (elect.) = 20.7% (
24、l-p.s.) = 15.7% (h-p.s.)930 kW(1.64 106Btu/h) = 36.2% (elect.) = 16.6% (hot.w) = 13.4% (h-p.s.)Boiler 11,280 kW 2(3.85 107Btu/h 2)7,520 kW 1(2.57 107Btu/h 1)( = 0.83)Not replacedChiller 1 Double-effect absorption chiller4,747 kW 4(1.62 107Btu/h 4)COP = 1.23Double-effect absorption chiller4,220 kW(1.
25、44 107Btu/h )COP = 1.51 (steam base)Chiller 2Not replacedChiller 3 Same as chiller 1Chiller 4 Not replacedChiller 5 Single- however, the differences in monthly boiler gas and accessory electricity consumption are larger than the others. This is because the measured electricity consumption includes e
26、lectrical equipment in the DHC plant, such as plant ventilation, which is not considered in the simulation. The base simulation considers only the cold-water pump, cooling water pump, cooling tower fan, absorption chiller, and boiler accessories. Figure 5 compares the chiller operational status betw
27、een the measured data, simulation with the baseline model, and simulation with the improved model. As shown in this figure, the number of chillers in actual operation is often smaller than in the baseline simulation result. The improvement of the model, as shown in following section, decreases the d
28、ifference between measured data and simulation results. 05,00010,00015,00020,00025,00030,00035,00040,00045,000Apr. May. Jun. Jul. Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar.Primaryenergyconsumption GJ/month04.79.514.219.023.728.433.237.942.7EnergyconsumptionGBtu/monthGas for Gas EngineGas for BoilerElec
29、tricity consumption for AccessoriesFigure 4 Comparison of monthly variation in energy consumption between measured and simulation results (From left to right, measured data, baseline simulation, Run-1, and Run-2 (i.e., improved simulation) 2011 ASHRAE 141MEASUREDGJ/YEAR (GBtu/YEAR) GJ/YEAR (GBtu/YEA
30、R) ERROR GJ/YEAR (GBtu/YEAR) ERRORGas for gas engine 80126 (75.9) 79478 (75.3) -0.8% 79478 (75.3) -0.8%Gas for boiler 205428 (194.7) 202174 (191.6) -1.6% 204008 (193.4) -0.7%Electricity for Accessories 51604 (48.9) 55285 (52.4) 7.1% 50907 (48.3) -1.4%Total 337158 (319.6) 336936 (319.3) -0.1% 334393
31、(316.9) -0.8%SIMULATION（before） SIMULATION（after）Table 2. Annual energy consumption 4/13/310:00 24:00No. of chillers 500RT*1 500RT*21350*01350*11350*21350*31350*4Improvement of simulation model Figure 5 Chiller operation status (From left to right, measured data, baseline simulation, and improved si
32、mulation) The following simulations are demonstrated step by step. Run-1: Degradation of the double-effect absorption chiller. Figure 6 compares the modeled and measured COP of the double-effect absorption chiller. The data was selected from two partial load ranges. The measured and simulated COPs d
33、o not agree well, especially for a low load factor, because of the difficulty of steam flow-rate measurement. However, the COP decreased for a high load factor, and therefore decreasing the COP for the 80100% partial load factor by 10% from the original value. As shown in Figure 4, the errors of Run
34、-1 became smaller than baseline during summer. 0.81.01.21.41.620 22 24 26 28 30 32 34COP-Cooling water Temp. Measured data Modeled characteristics0.81.01.21.41.620 22 24 26 28 30 32 34COP-Cooling water temp Measured data Modeled characteristics(68) (71.6) (75.2) (78.8) (82.4) (86) (89.6) (93.2) (68)
35、 (71.6) (75.2) (78.8) (82.4) (86) (89.6) (93.2) (F) (F) Figure 6 Comparison between modeled COP and measured value. (Left: Load factor 90%100%, Right: Load factor 60%70%)Run-2: Change the dead band margin and minimum flow rate of the bypass. The dead band margin is set to 10% and the minimum flow ra
36、te of the bypass is set to 0 m3/h (0 ft3/h), because the number of chillers in actual operation is often smaller than the simulated result. As shown in Figure 4, the errors in the amount of gas for boilers reduced. Simulation result with the improved model (i.e., Run-2) As shown in Table 2, the diff
37、erence between the measured values and improved simulation results for annual measurements is smaller than that before the improvement. As shown in Figure 5, the improved simulation results came closer to the measured condition. However, small differences still exist at midnight. At this time, under
38、 actual conditions, 142 ASHRAE Transactionsonly one small single-double effect absorption chiller is in operation, even if the flow rate and cooling load exceed the rated capacity. This is because the chiller capacity increases as the cooling water temperature decreases. This energy saving operation
39、 is based on decisions made by an experienced operator, and is not considered in our model. SETTING PARAMETERS WITH THE NEW SYSTEM SIMULATION In this chapter, the post-renovation simulation model was developed based on the simulation model described in the previous section. Modeling equipment perfor
40、mance The renovated plant has one gas-engine CHP system. It supplied electricity and waste heat, which consists of high-pressure (0.78 MPa (7.8 bar) steam and hot water (88C (190.4F), to the DHC plant. It operates constantly at 100%. Figures 7 and 8 show the COP characteristics of a variable speed t
41、urbo chiller and waste-heat utilization absorption chiller installed during renovation. Energy consumption of variable speed pump was modeled as a function of the flow rate from the catalog. (68F) (75.2F) (82.4F) (89.6F)0246810204060801COP-Load factor%(68F) (75.2F) (82.4F) (89.6F)05101520250 20 40 6
42、0 80 100 120COP-Load factor%12(53.6F)13(55.4F)15(59F)20(68F)25(77F)30(86F)32(89.6F)Cooling water enterning temperature020 24 28 32 32 28 24 201.0 1.2 1.4 1.6 1.8 2.0 0 2040608010Load Factor%COP-Figure 8 COP characteristics of waste-heat utilization absorption chiller （ Left: using waste heat, Right:
43、 without using waste heat） Figure 7 COP characteristics of variable speed turbo chiller 1/10:00 24:001/31Modeling of chiller control sequence After the renovation, the operation schedule differs according to seasons and the switching ON/OFF of the CHP. The chiller operating order was planned to swit
44、ch among three patterns: 1. When the gas engine is in operation (daytime on weekdays and saturday) Chiller 5 6 1, 3 2, 4 2. When the gas engine is suspended Chiller 6 1, 3, 5 2, 4 3. Daytime on weekdays in July and August (peak-cut mode) Chiller 5 1, 3 2, 4 6 *Waste-heat utilization absorption chill
45、er must consume waste hot Turbo chiller absorption chiller1Turbo + absorption chiller1 absorption chiller2Turbo + absorption chiller2 absorption chiller3Turbo + absorption chiller3 absorption chiller4Turbo + absorption chiller4 absorption chiller5Turbo + absorption chiller5water when the CHP operate
46、s. SIMULATION OF THE RENOVATED PLANT Figure 9 Chiller operation status (From left to right, measured data, baseline simulation, and improved simulation) Simulation result by baseline model To determine the parameters used in the simulation model, actual operating data from 2009 was compared with sim
47、ulated results for the plant after the renovation. Figure 8 shows the comparison of chiller operation status between the data, 2011 ASHRAE 143simulation with the baseline model, and simulation with the improved model. As shown in Figure 9, the numbers of chillers in actual operation and in the basel
48、ine simulation are almost the same during winter; however, for the rest of the year, the number of chillers in the baseline simulation is smaller than that for actual operation. Figures 10, 11, and Table 3, respectively, show the comparisons of the hourly, monthly, and annual measured energy consump
49、tion, baseline simulation, and improved simulation results. The annual energy consumption of the baseline simulation result shows good agreement with the measured values; for example, the annual error is 3.5%. However, the hourly energy consumption of the baseline simulation is smaller than the measured values through the entire day. Furthermore, the monthly energy consumption of the baseline simulation is smaller 10152025301:003:005:007:009:0011:0013:
