Use EMS to Improve Simulation of Outside Air Economizer and Fan Control for Unitary Air Conditioners

Transcription

1 Building Energy Simulation Forum Use EMS to Improve Simulation of Outside Air Economizer and Fan Control for Unitary Air Conditioners Reid Hart, PE with help from: Rahul Athalye; Weimin Wang, Ph.D.; Michael Rosenberg January 2014 PNNL-SA

2 Outline RTUs, Economizers & Fans OSA Economizer Modeling Energy Management System (EMS)* in EnergyPlus Economizer simulation with EMS Fan speed reduction simulation with EMS Overall energy impact *EMS here refers to an EnergyPlus feature that provides a way to develop custom control and modeling routines for EnergyPlus models. 2

3 RTUs: Economizers & Fans

4 Why Rooftop Units 4 Significant Volume Represents over 40% of commercial building area. Energy Code Opportunities DCV on RTUs serving high occupancy Single Zone VAV on RTUs > 9 tons Economizer performance (integrated vs. non integrated) 4

5 Why HVAC 5 Because HVAC energy represents a significant percentage of energy use in a commercial building ~ 60% Largest percentage of savings is gas Misc. Equip. 24% Typical Office Building Area Lights 19% Space Cool 8% Vent. Fans 14% Space Heat 35%

6 Breakdown of HVAC Energy 0.4 Office (Functioning Economizer) Space Cool (kbtu/sf) Vent Cool (kbtu/sf) Vent Heat (kbtu/sf) Space Heat (kbtu/sf) Fans (kbtu/sf) Econo Savings (kbtu/sf) Energy Use Breakdown Cooling = 15% / Heating = 55% / Fan = 30% 6

7 Basic RTU Components T Damper Actuator Digital Controller VSD or 2 spd CO2 Sensor OA Temp Fan Airflow Control: 33% Fan Energy Savings Ventilation Control: 50% Ventilation Energy Savings Economizer Control: 33% Cooling Energy Savings Ventilation Cooling Economizer Cooling Ventilation Heating 7 Space Cooling Fan Energy Space Heating

8 RTU Cooling Energy Use 8 Million Btu's Baseline HVAC Energy Use Fan Heat Cool Cooling is common down to 44 F OAT in offices in winter OSA Te m perature Bin 8 No Economizer

9 Economizer Energy Savings Potential 9 Million Btu's ECM HVAC Energy Use Fan Heat Cool % Cooling Energy Savings OSA Temperature Bin 9 Optimized Economizer

10 Variation of Supply Fan Airflow 90.1 addendum AQ added fan control requirements phasing in 110 MBH 2013; 75 MBH 2014; 65 MBH 2016 Source: Carrier 10

11 Need for Improved Simulation Addendum AQ to introduced several items: Multi-stage or variable cooling capacity to improve economizer and fan operation. Economizers on RTUs with direct expansion (DX) mechanical cooling to remain fully open during integrated operation until discharge air temperature (DAT) drops to 45 F (7 C). Multi-speed or variable speed fans to reduce fan energy use. Future items under consideration Reduced return (RA) air damper leakage Current modeling would not show savings: Economizer operation is now modeled as ideal Multi-speed fan not correctly implemented in EnergyPlus 11

12 PNNL Prototypes for 90.1 ASHRAE 90.1 progress indicator: Tracks progress between versions of ASHRAE Standard 90.1 Results are weighted for national construction occurrence of the prototypes E+ files available (search: PNNL prototypes) Stand Alone Retail Prototype: Used for this example Single floor 24,695 square feet construction 12 Package Rooftop Units (RTU) with dry-bulb air economizer

13 OSA Economizer Modeling

14 RTU Economizers: Modeling vs. Reality Modeling of unitary system economizers in EnergyPlus & DOE2 Modeled in a single mode per time step; actually multiple modes Assume full economizer benefit is available, regardless of DAT Operational reality When economizer operates alone, full economizer is available With full economizer, typical outside air fraction is closer to 70% In integrated operation, low DAT closes economizer 14 A m ps / Volts Tem perature, o F :52 AM 10:12 AM 10:32 AM 10:52 AM 11:12 AM 11:32 AM 11:52 AM AMPS TSTAT OSA DA RA

15 DX Economizer Alternating Integration Alternating integration means 2 cooling modes Full economizer without DX cooling coil operation Partial economizer with DX cooling coil operation needed to: Maintain comfortable discharge air temperature Avoid coil freezing Source: PECI & Taylor Engineering,

16 Economizer Effectiveness Ideal OSA economizer contribution to sensible cooling (100% OA) (RA-OA) [time] Is impacted by the actual % of OA (RA-OA) [time] [%OA] We need to cover both alternating modes for DX cooling (RA-OA) [t 1 ] [%OA withdx ] + (RA-OA) [t 2 ] [%OA nodx ] e.g., (75-65) [t 1 ] [40%] + (75-65) [t 2 ] [70%] By magic of algebra, for t = timestep (RA-OA) t ( [%t 1 ] [%OA withdx ] + [1-%t 1 ] [%OA nodx ] ) or, (RA-OA) t [economizer effectiveness %] where [%t 1 ] = f (Load%, OA DB, OA WB ) Economizer effectiveness for time step is input into E+ with EMS using the economizer effectiveness in a schedule as the maximum outdoor air fraction in the outdoor air controller object 16

17 Maximum Economizer Outside Air Fraction Smaller RTUs have lower cost economizer dampers High quantity of return air leakage Acknowledging this impact allows future credit for improved return air damper seals Source: Davis,

18 Find Economizer Effectiveness Mostly, cooling control is a function of dry-bulb space temperature So if we correctly determine the sensible load contribution of the economizer, our input will be correct, the model will adapt DX cooling capacity is strongly impacted by entering wet-bulb So, to find [economizer effectiveness %], we need to solve [%OA withdx ] + [1-%t 1 ] [%OA nodx ] )where [%t 1 ] = f (Load%, OA DB, OA WB ) So run many humidity conditions to find eff ECON given know DX operation Matrix eff ECON as f (Load%, OA DB ) for multiple OA WB & RA WB 18

19 Economizer Contribution Relates to OAT Maximum %OA during DX cooling is limited by comfort Economizer-only time is relative to internal load Share of ideal economizer depends on OAT & load Source: PECI & Taylor Engineering,

20 Simplified Economizer Effectiveness For a range of loads, time in each stage of cooling found Impact of alternating integration on economizer performance calculated, assuming 70% maximum With more stages of DX, impact not as great 20

21 EMS Modeling in EnergyPlus

22 Custom Simulation Control 22 equest / DOE2 Two different methods Input Macros (v2.2) allow parts of BDL files to be mixed and matched Selective skipping parts of input file before run Defining blocks for reference Apply math to input generation User Functions Allow on the fly runtime control changes Limited to versions 2.1e May be available in the 2.2 engine from text BDL, but not recognized or editable by the equest database EnergyPlus E+ also has input macros EMS = Energy Management System Erl = E+ Runtime Language Intended to simulate functions normally available in an EMS Can be used for other conditional inputs during run Essentially adjusts Actuators based on Sensors EMS Examples: Reset flows based on load or environmental conditions Activate humidification only under certain conditions

23 Economizer in Erl EMS Erl code requires: 1. Sensors (inputs) 2. Internal Variables 3. Actuators (outputs) 4. Global Variables 5. Programs 6. Program Calling Managers See the separate Application Guide for EMS (a.k.a. The Book of Erl) in the E+ documentation 23

24 Why do we need to use E+ EMS? Why not just use short time steps? Then E+ can alternate between economizer and not economizer, right? No it won t E+ does not model true temperature differential control (nor does DOE2) For cycling units, the capacity & load is needed vs. a power curve that has cycling losses built in. Ppwer is used for power for the whole time step! Also, going from 15 to 1 minute timestep = long run times (260 runs/ecm) Since only one mode is modeled we need to fool E+ by Entering a maximum OA flow that matches the average OA airflow during the time step Entering fan power equivalent to the actual average fan power during the time step 24

25 Economizer Logical Steps in Erl Establish Stepwise logic 1. Coil Entering Conditions (MAT) when OAT at minimum ventilation position 2. Coil Entering Conditions (MAT) when OAT at maximum economizer position 3. Find MAT at each stage of cooling (may need iteration) 4. Find total sensible dt at each stage, including economizer benefit dt 5. Find stages operating during the time step & MAT each stage 6. Find economizer effectiveness for each stage and average during the time step 7. Find Fan average kw during the time step 25

26 Economizer Logical Steps in Erl 26

27 Abandoned full Erl Implementation Abandoned full Erl Implementation Difficult to debug Long runtime, as MAT was iterative calculation Perform step by step calculations in a spreadsheet for range of conditions Max OA Fraction is f(cool Load, OAT WB, OAT DB ) Develop from regressions 27

28 Streamline Code with Regressions 28

29 Note on EMS Algebra & Quality Assurance Use more parenthesis than you ever thought you would need! EMS/Erl does not understand algebraic priorities!!! Example: economizer effectiveness regression for 2- & 1-stage cooling: As always, inspect hourly or time step runs with output of Sensor inputs; Intermediate variables; Actuator outputs Once debugged can comment out to improve speed 29

30 Regression Development Full simulation with EMS resulted in long run times DX coil capacity adjustment determined based on entering temperature, dry bulb and wet bulb The required DX cooling run time was determined At multiple loads For multiple outside dry bulb (DB) and wet bulb (WB) conditions For multiple return air wet bulb conditions Economizer operation in both modes determined 3 minute delay for addendum AQ 45 F integration requirement Effective maximum economizer operation determined Regressions for 1-stage (R ) and 2-stage (R ) Regressions programmed into EMS for max OA fraction Based on cooling load, OA WB, OA DB (RA WB not significant) 30

31 Regression: Economizer Effectiveness OAT, DB, F: 72 F 65 F 58 F Two-Stage Single Stage 31

32 Energy Impact of Improved Model Improved model reduces economizer effectiveness and more correctly models mechanical cooling energy use 32

33 Fan Modeling

34 RTU Fan Speed Control Small RTUs use multi-speed fans, not variable speed drives (VSDs) By affinity laws, fan power will reduce as some exponent of airflow However the modeled operation does not match time-weighted power Example: 30% coil load has 30% time at full speed with the DX coil on the rest at 50% speed for ventilation or economizer 34

35 Time vs. Flow Weighting EnergyPlus and other models use flow weighting for fan power on unitary systems EnergyPlus allows different fan speeds for cooling, heating, ventilation, but not economizer mode Sequence for first stage cooling may have: Low speed when economizer locked out Full speed when economizer is OK Low speed for partial economizer Actual operation is time weighted; 100% cooling, 50% ventilation Flow-weighting estimates fan power 20% low at 30% cooling load Actual RTU Operation Time-weighted Flow-weighted Flow vs. Factor High Speed Low Speed Average Average Time Weighting Mode Cooling Ventilation Mixed Time in Mode 30% 70% Mass Flow or CFM (Q) 100% 50% 65% Fan Power Ratio (W ratio ) % Low 35

36 EMS Fan Power Development EMS calculations in each time step Use the coil fraction to determine time in each mode Adjust speed for economizer operation Calculate the time-weighted fan power for time step Fan power is proportional to static pressure, so Set the fan in E+ for full speed/ full flow Apply an adjustment to the static pressure The proper fan power is calculated Fan energy matched the multi-speed fan use by mode Fan heat added to airstream is correct for time step 36

37 Determine Fan Control in each Time Step 37

38 Other Examples of Use of EMS Turn on a relief fan when economizer is in use Simulate pressure control If OA > Min Ventilation Energize relief fan Reduce cooling temperature setpoint if space humidity exceeds threshold If RH > 60% Cooling setpoint deg. F 38

39 Overall Results

40 HVAC Energy End Use Salem Oregon, climate zone 4C Stand-Alone Retail prototype Correct for less economizer cooling, then improvements from two-stage cooling, then corrected multi-speed fan operation savings 40

41 HVAC Energy End Use HVAC Energy Use Index 1000 Btu s per square foot Stand-Alone Retail prototype Overall impact in multiple climate zones 41

42 Conclusions EnergyPlus (and DOE2) understate RTU energy use Economizer effectiveness is overstated Fan savings from speed reduction is overstated In EnergyPlus EMS can be used to improve results Based on cooling load, OA WB & OA DB, economizer effectiveness is found in each time step Fan time-weighted power is found in each time step 42 Future possibilities Revise RTU native simulation in programs to account for multiple modes occurring in one time step Determine time in each mode to meet load Calculate each mode operation and energy impact separately Sum mode results for the time step

43 References ASHRAE. (2010). ASHRAE/ASHRAE/IES Standard : Energy Standard for Buildings except Low-Rise Residential Buildings. American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc., Atlanta, GA. Davis, R., Francisco, P., Kennedy, M., Baylon, D., and Manclark, B. (2002). Enhanced Operations & Maintenance Procedures for Small Packaged Rooftop HVAC Systems: Protocol Development, Field Review, and Measure Assessment. Ecotope for Eugene Water & Electric Board. EIA. (2008) Commercial Buildings Energy Consumption Survey. U.S. Energy Information Administration. Hart, R. (2011). DOE Work-Arounds to Hidden Problems: equest - A half hour to learn; three years to master. Portland, OR. Hart, R., Price, W., and Morehouse, D. (2006). The Premium Economizer: An Idea Whose Time Has Come. Proceedings of the 2006 ACEEE Summer Study on Energy Efficiency in Buildings, American Council for an Energy-Efficient Economy (ACEEE), Pacific Grove, CA, LBNL. (2012a). EnergyPlus Engineering Reference: The Reference to EnergyPlus Calculations. Lawrence Berkeley National Laboratory for USDOE. LBNL. (2012b). Application Guide for EMS: Energy Management System User Guide. Lawrence Berkeley National Laboratory for USDOE. PECI & Taylor Engineering. (2011) Light Commercial Unitary HVAC for Codes and Standards Enhancement Initiative (CASE). California Utilities Statewide Codes and Standards Team for California Public Utilities Commission. Thornton, B. A., Wang, W., Cho, H., Xie, Y., Mendon, V. V., Richman, E. E., Zhang, J., Athalye, R. A., Rosenberg, M. I., and Liu, B. (2011). Achieving the 30% Goal: Energy and Cost Saving Analysis of ASHRAE/IES Standard PNNL for USDOE. 43

44 Questions? Reid Hart, PE