Report

btrc.masdar.ac.ae Optimal Control of Chiller Condenser Sub-cooling, Compressor Speed, Tower Fan and Pump Speeds, and IGV Omer Qureshi, Hassan Javed & Peter Armstrong, June 2013 1 Presentation Outline Introduction SCADA and Heat Balance Analysis Component Models Chiller System Solver Optimization Conclusion and Future Work 2 Introduction Plant under consideration-(4x2500T). Collection and analysis SCADA Development of sub models for Individual chiller components Validation of model Development of solver- to execute these sub models and predict chiller performance. Optimize the model to produce set of conditions for optimum power consumption. 3 District Cooling Plant Selected District cool Plant Capacity (4x2500T) Shell and tube Evaporator and Condenser Constant speed single stage centrifugal compressor Capacity control by Pre-rotation vanes Surge control Variable geometry diffuser 2-cell cooling tower each with variable speed fan (Fan diameter: 8m) Variable speed chilled water pump Constant speed condenser water pump 4 Chiller Unit 1. Maintenance manual of York Chiller(Source: Tabreed) 5 SCADA & Heat Balance Analysis 6 Components Models—Chiller Unit Steady-state models based on first principle Inputs Component engineering parameters SCADA Data Simple models, less computation time Four Component models for district cooling plant Evaporator Model----Shell and tube Condenser Model----Shell and tube Centrifugal Compressor Model (Isentropic work + loss Mechanism) • Constant speed • Variable speed Variable speed pump model 7 Evaporator Model ENGINEERING PARAMETERS Tubes Length of shell Copper 6.6 m Tube Pass (water) 2 Total no. of tubes 1234 Tube Diameter 0.75" or 1.905x10-2 m Tube thickness 0.028" or 7.11x10-4 m Assumptions: No pressure drop considered for refrigerant side Thermal resistance from refrigerant side is neglected. 8 Evaporator Model , , ,, Evaporator ,,1 Evaporation ,,2 Evaporation ,,2 ,, Superheating . 1st Pass 2nd Pass Two regions for refrigerant were modeled: Evaporation Superheating – NTU Method Single Stream HX for evaporation Crossflow HX for super heating 9 Evaporator Model Equations utilized in Evaporator Model ,, = ,, − ℎ, = 0.023 0.8 0.4 , , = , = 1 1 , ℎ, + , min[, , , , ] ,,1 = , ( /2) ,,1 = , ( /2) ,,1 = , (1 − ) ( /2) Evaporation Evaporation Superheating 10 Evaporator Model Equations utilized in Evaporator Model Equation for regressed length: = 8.94710−3 2 − 3.627910−1 + 7.227 Equation for temperatures: ,,1 = ,, − (,, − ) 1 − −1 ,,2 = ,,1 − (,,1 − ) 1 − −,2 ,,2 = ,,2 − 2 (,,2 − ) 2 = 1 − 1 2 0.22 exp − ,2 0.78 −1 11 Evaporator Model 1. Maintenance manual of York Chiller(Source: Tabreed) 12 Evaporator Model 4 0.2096 C NRMS 0.0319 3.5 Modeled Te (C) RMS Measured Te (C) vs Modeled Te (C) Measured Te (C) 15% error line -15% error line 3 2.5 2 1.5 1.5 2 2.5 3 Measured Te (C) 3.5 4 13 Condenser Model ENGINEERING PARAMETERS Tubes Length of shell Tube Pass (water) Total no. of tubes Sub-cooling Section: Tube Diameter No. of tubes Tube thickness Tube Surface Area Condensation & de-superheating Section: Tube Diameter No. of tubes Tube thickness Tube Surface Area Copper 6.6 m 2 937 0.75" or 1.905x10-2 m 180 0.028" or 7.11x10-4 m 66.78 m2 1" or 2.54x10-2 m 757 0.035" or 8.89x10-4 m 376.44 m2 Assumptions: No pressure drop considered for refrigerant side Thermal resistance from refrigerant side is neglected. 14 Condenser Model Condenser ,, , . ,, ,, Condensation ,,2 Condensation ,,2 Desuperheating Sub-cooling 1st Pass 2nd Pass Three regions for refrigerant were modeled: Sub-cooling Condensation De-Superheating – NTU Method 15 Condenser Model Equations utilized in Condenser Model 1a. Sub-Cooling Section(First Pass): ℎ,,1 = 0.023 1 0.8 0.4 ,1 = 1 1 ,,1 ℎ,,1 + ,,1 = 2 − ,,1 ,1 = 1, 1 − −,1 1−,1 1 − ,1 −,1 (1−,1) ,1 ,1 (2 − ,, ) , , (2 − ) = ,, + , 1, , 16 Condenser Model 1b. Condensation Section (First Pass): ℎ,,1 = 0.023 1 0.8 0.4 ,1 = 1 1 ,,1 ℎ,,1 + ,,1 ,,1 2, ,1 = ,1 , (2 − 3 ) = ,, + , (1 − 1 ), Mixing Section: ,, = , 1, ,1, − , (1 − 1 ) ,1 , 17 Condenser Model 2a. Condensation Section (Second Pass): ℎ,,2 = 0.023 2 0.8 0.4 ,2 = 1 1 ,,2 ℎ,,2 + ,,2 ,,2 = ,, + 2, ,2 = ,2 , , (2 − 3 ) , , 2b. De-superheating Section (Second Pass): ,,2 = ,2 + ,,2 (1 − 2 ) , , 18 Condenser Model 0.0949 C NRMS 0.0225 Measured Tc (C) vs Modeled Tc (C) Measured Tc (C) 2.5% error line -2.5% error line 32 30 Modeled Tc (C) RMS 34 28 26 24 22 22 24 26 28 30 Measured Tc (C) 32 34 19 Condenser Model 35 0.6481 C NRMS 0.1471 30 Modeled Tw,out (C) RMS Measured Tw,out (C) vs Modeled Tw,out (C) Measured Tw,out (C) 5% error line -5% error line 25 20 20 25 30 35 Measured Tw.out (C) 20 Compressor Model Integral and mathematically most complex part of chiller Constant and variable speed compressor model Non-Dimensional loss model based on Aungier(2000) Assumptions Centrifugal Compressor Specification • Gear efficiency is taken as 90% Refrigerant • Velocity profile is assumed as constant, along the hub and tip Rating (Btuh) 2500 • The kinetic energy of refrigerant entering the diffuser is completely converted to useful energy Rating (kW input) 1817 Rating discharge pressure (psig) 162 • Diffuser and IGV losses are not modeled Rating suction pressure psig) 34 • Water flow rate for motor cooling is taken as constant • Complex engineering parameters in impeller geometry Rating suction temperature (F) R134A 33/34 Impeller diameter (outlet diameter) m 0.7 Impeller hub diameter (inlet diameter) 0.3 Impeller Blade Angle (degree) 45/50 21 Compressor Model-Inputs and Outputs Constant Speed Model Input IGV Positions Constant RPM Inlet and outlet pressure of compressor Inlet and outlet blade and velocity angles of impeller Impeller Inlet and outlet engineering parameters and dimensions Gear efficiency Output Compressor Power Pressure at impeller exit Temperature at compressor outlet Pressure drop due to Impeller losses Variable speed Model Input Mass flow rate of refrigerant Inlet and outlet pressure of compressor Inlet and outlet blade and velocity angles of impeller Impeller Inlet and outlet engineering parameters and dimensions Gear efficiency Output Compressor Power Compressor RPM Pressure at impeller exit Temperature at compressor outlet Pressure drop due to Impeller losses 22 Validation Constant Speed Compressor Model 1600 Actual Power(kW) Model Power(kW) Loss Power(kW) Model Comp Power(kW) 1400 Copmressor Power (KW) 1200 1000 800 600 400 200 0 0 200 400 600 800 No. of Observations 1000 1200 1400 23 Validation Constant Speed Compressor Model 1600 108.34 KW NRMS 0.1553 1400 1200 Model Power(kW) RMS Measured Power(kW) vs Model Power(kW) Measured Power(kW) 10% Error line -10% Error line 1000 800 600 400 400 600 800 1000 1200 Measured Power(kW) 1400 1600 24 Variable Speed Compressor Model = 1 = − 1 1 −1 3 1 −1 = = 2 2 2 2 RPM is calculated in an iterative process by satisfying the following equation = − Total Work = + Loss Model Calculations = ∆ Total Relative Pressure Drop (Due to Losses) ∆ = ( 1 − 1 ) 25 Variable Speed Compressor Model-Benefits/comparison Compressor Power (KW) Variable Speed Compressor (KW) Measured Compressor Power (KW) Power (KW) IGV Position 1504.702 44.2 Operation Conditions: 1. mr (kg/s) 2. Pout/Pin No. of Observations 26 Impeller Loss Model 1 = 1 − 1 sin( 1 ) 1ℎ = 0.8 1 − 1 2 1 + 2 1 sin( 1 ) 2 2 − = 4 1 2 (∆ 1 )2 = 24 ( − 1) 2 = 1 = 2 2 ∆ 1 12 2 ( 1 ) − ℎ = 6 27 Variable Speed Compressor Model-losses profile 120 Pressure Drop (kPa) 100 80 Clearance gap loss (kPa) Diffusion loss (kPa) Hub-shroud Loss (kPa) Incident loss (kPa) Skin friction loss (kPa) Blade Loading Loss (kPa) Expansion Loss (kPa) 60 40 20 0 20 25 30 35 Refrigerant Mass Flow (kg/s) 40 45 50 28 Cooling Tower Model Effectiveness NTU Method = = ∗ ∗ ( − ) = − ∗ = ∗ ∗ ( − ) 1 − −(1−) = 1 − −(1−) = _ _ ∗ ∗ = _ Regression Coefficient = _ = ∗ ∗ 29 Cooling Tower Model Assumptions, Specifications and Input/ Output Variables Assumptions • • Cooling Tower Specifications Air exiting the tower is saturated with water Rating (RT) 5000 vapor and is only characterized by its Rating flow rate (GPM) 15300 enthalpy Rating ambient wet bulb (F) 86 Rating ambient dry bulb (F) 122 Rating entering condenser water 105 Reduction of water flow rate by evaporation is neglected in the energy balance. • Mass flow rate is calculated by considering linear proportionality of mass flow rate of air and motor speed. temperature (F) Fan diameter and speed (m, RPM) 8/152.6 Air flow rate (CFM) 776383 Inputs • • • • • Wet-bulb temperature Cooling tower supply water temperature Dry-bulb temperature Mass flow rate of water Cooling tower fan/motor speed Outputs • • Cooling tower return water temperature Merkel’s Number 30 Cooling Tower Model 31 Pump Model Mainly there are two mode of operation for these pumps: Constant flow pump Variable flow pump with a variable speed drive To model a variable pump power following relationship is used: = (1 + (2 ) + (3 )2 +(4 )3 ) Where, PMP = pump motor power at rated condition, kW C1, C2, C3 and C4 are pump performance coefficients Also, PLRi = pump part load ratio defined as follows: = = 32 Pump Model Validation Graph + 5%Error Line 33 Solver Description Qt,e Tw,in,e Tw,in,c Ve Vc dTsh,e 34 Optimization Optimization performed with two configurations: Chiller Water Flow Optimization Chiller Water Flow And Condenser Water Flow Optimization Objective Function: Minimize total power consumption i.e. compressor power and pump(s) power combined. 35 Optimization Chiller Water Flow Optimization: Tw,in,c = 25 C and Tw,in,e = 14 C Vc Qe 0.4795 m3/s 10000 KW Power Ve Total (KW) (m3/s) 2791.90 2494.66 2325.43 2226.70 2171.34 2145.79 2149.01 2177.04 2227.85 2300.43 2389.48 0.1419 0.1774 0.2129 0.2484 0.2839 0.3194 0.3548 0.3903 0.4258 0.4613 0.4968 COP 3.58 4.01 4.30 4.49 4.61 4.66 4.65 4.59 4.49 4.35 4.19 Vc Qe 0.4795 m3/s 8000 KW Power Ve Total (KW) (m3/s) 1768.81 1617.53 1535.14 1492.30 1476.36 1483.94 1512.07 1559.59 1623.07 1708.17 1809.82 0.1419 0.1774 0.2129 0.2484 0.2839 0.3194 0.3548 0.3903 0.4258 0.4613 0.4968 COP 4.52 4.95 5.21 5.36 5.42 5.39 5.29 5.13 4.93 4.68 4.42 Vc Qe 0.4795 m3/s 6000 KW Power Ve Total (KW) (m3/s) 1102.50 1032.45 997.69 988.83 998.06 1023.13 1065.53 1122.92 1197.94 1288.91 1397.89 0.1419 0.1774 0.2129 0.2484 0.2839 0.3194 0.3548 0.3903 0.4258 0.4613 0.4968 COP 5.44 5.81 6.01 6.07 6.01 5.86 5.63 5.34 5.01 4.66 4.29 Vc Qe 0.4795 m3/s 4000 KW Power Ve Total (KW) (m3/s) 649.15 626.16 622.12 633.21 657.75 695.26 746.65 811.97 892.40 988.98 1103.05 0.1419 0.1774 0.2129 0.2484 0.2839 0.3194 0.3548 0.3903 0.4258 0.4613 0.4968 COP 6.16 6.39 6.43 6.32 6.08 5.75 5.36 4.93 4.48 4.04 3.63 36 Optimization Total Power (KW) Chiller Water Flow And Condenser Water Flow Optimization: Qe Ve,opt Vc,opt = 10,000 kW = 0.349 m3/s = 0.408 m3/s Tw,in,e = 14 C; Tw,in,c = 25 C 37 Optimization Total Power (KW) Chiller Water Flow And Condenser Water Flow Optimization: Qe Ve,opt Vc,opt = 8,000 kW = 0.296 m3/s = 0.355 m3/s Tw,in,e = 14 C; Tw,in,c = 25 C 38 Optimization Total Power (KW) Chiller Water Flow And Condenser Water Flow Optimization: Qe Ve,opt Vc,opt = 6,000 kW = 0.249 m3/s = 0.332 m3/s Tw,in,e = 14 C; Tw,in,c = 25 C 39 Optimization Total Power (KW) Chiller Water Flow And Condenser Water Flow Optimization: Qe = 4,000 kW Ve,opt = 0.205 m3/s Vc,opt = 0.251 m3/s Tw,in,e = 14 C; Tw,in,c = 25 C 40 Optimization Chiller Water Flow And Condenser Water Flow Optimization: Tw,in,e = 14 C; Tw,in,c = 25 C 41 Optimization Chiller Water Flow And Condenser Water Flow Optimization: 42 Conclusions Variable Speed compressor provide savings of 30-40% Variable speed pump for water circulation play an imperative role in reducing overall power consumption of chiller plant. Modeling of chiller components can be performed with limited engineering information from manufactures. 43 Future Work More rigorous compressor loss model Transient model for the condenser and evaporator Cooling tower Model Variable Speed condenser pump Investigate the effect of pressure drop and resistance from refrigerant side 44 Q&A 45