Tracking the Benefits of Retro-Commissioning - Building ...

6 downloads 17069 Views 908KB Size Report
May 2, 2007 - Matt Denny. Quantum Energy Services & Technologies, Inc. Raul Abesamis. Physical Plant-Campus Services. University of California, Berkeley.
National Conference on Building Commissioning: May 2-4, 2007

Tracking the Benefits of Retro-Commissioning: M&V Results from Two Buildings David Jump, Ph.D., P.E. Matt Denny Quantum Energy Services & Technologies, Inc. Raul Abesamis Physical Plant-Campus Services University of California, Berkeley Synopsis This paper will present measurement and verification (M&V) results of energy savings from commissioning projects at two of the University of California at Berkeley’s buildings: Soda Hall, a computer-science building, and Tan Hall, a chemistry building. The M&V process is an independent methodology used to validate savings estimated for system tune-ups that were identified during the commissioning process. The verified savings illustrate how well savings were estimated using the standard techniques of bin methods and computer simulations. These projects are part of the University of California/California State University’s MonitoringBased Commissioning Program. This paper summarizes the M&V methodology (originally presented at the 2005 NCBC) and discusses the advantages of integrating M&V into commissioning projects. These advantages include: providing operators with tools to maintain and improve energy performance, accurate accounting of savings over time, diagnostic abilities of the method, obtaining LEED-EB credits, improving the persistence of savings, and improving energy efficiency program reporting and third party review processes. Practical considerations in implementing an M&V program are discussed, such as complying with industry-standard M&V guidelines, identifying and isolating systems, obtaining reliable data, and programming requirements. The strategies used to minimize the costs of adding points are discussed. The procedures required for establishing the energy baseline, tracking uncertainty, and estimating energy savings are presented. The discussion also includes how these techniques may be included in developing performance monitoring specifications for commercial building control systems, and how the methods may be incorporated into the routine activities of the operations staff. About the Authors David Jump is Quantum Energy Services & Technologies’ (QuEST) Director of Engineering. He currently leads a team of eleven engineers in providing commissioning and energy project services at QuEST. He has developed and implemented public-goods funded retroJump et al: Tracking the Benefits of RCx.

1

National Conference on Building Commissioning: May 2-4, 2007

commissioning programs since 1999 for the City of Oakland, and in utility service territories throughout California. He has authored chapters in federal and utility program M&V guidelines, and developed tools for assessing measurement protocol cost effectiveness. He is the current chair of the Efficiency Valuation Organization’s Revisions Subcommittee, which will publish the latest revision of the industry standard measurement and verification guideline, the IPMVP, this year. He is also the current vice-president of the Southwest Chapter of the Building Commissioning Association. He received a Ph.D. in Mechanical Engineering from the University of California, Santa Barbara and has over 15 years experience in the energyengineering field. Matt Denny is an Engineer at QuEST. He has a B.S. in Mechanical Engineering from San Francisco State University. Mr. Denny is lead engineer on QuEST’s monitoring-based commissioning projects with the University of California. Mr. Denny has recently completed retro-commissioning projects for UC Berkeley, Oracle Corporation, and Lockheed-Martin. He is expert with multiple data collection systems and analysis tools, energy analysis techniques, and optimizing building systems operation. Raul Abesamis has a B.S. in Industrial Engineering and is currently an Energy Engineer at the University of California at Berkeley Physical Plant-Campus Services department. He has developed and implemented energy conservation projects at the University with focus on improving energy efficiency through lighting retrofits and HVAC and controls systems automation. Mr. Abesamis is the campus project manager for the implementation of the monitoring based commissioning and retrofits programs under the UC/CSU/IOU Partnership Program. Acknowledgements The authors would like to thank Karl Brown, Deputy Director of the University of California’s California Institute for Energy and the Environment, and Michael Anderson, P.E. Principal of Newcomb Anderson McCormick for insightful comments and cost data received in the preparation of this paper.

Jump et al: Tracking the Benefits of RCx.

2

National Conference on Building Commissioning: May 2-4, 2007

Introduction Retro-commissioning (RCx) processes involve the collection of operational data and the performance of functional testing of building systems to identify and correct flaws in operations and optimize performance to meet the owner’s requirements. Depending on the project scope, recommendations are made and implemented to improve energy efficiency, thermal comfort, air quality, and other building services. Energy efficiency programs that offer RCx services identify and implement low-cost tune-up and optimization measures, that typically reduce a building’s energy use by 5 to 15%1. Most RCx projects recommend measures based on estimated energy savings and costeffectiveness. Savings estimates are made using various techniques such as engineering and statistical modeling, bin analysis, or available simulation programs for whole-building energy use. The level of detail and accuracy in savings estimates varies widely based on project scopes and budgets, and goals of the involved parties. Many RCx measures, such as simple set point or schedule changes, have short lifetimes because they are easily defeated. An owner or a publicgoods program manager sponsoring an RCx investigation and implementing changes desires to have confidence in the savings, and to extend the benefits of this investment for as long as possible. In order to assure that these savings have been achieved, and that they will endure, there is a clear need to effectively measure and verify the savings. Established measurement and verification (M&V) guidelines, such as the International Performance Measurement and Verification Protocol2 (IPMVP), and ASHRAE’s Guideline 143 provide an appropriate framework that overlaps with many RCx standard practices. These M&V methods are designed to provide independent verification of any savings estimation, and produce reliable results with methods that can be repeated by other parties. The California Public Utilities Commission, with its adoption of the Evaluator’s Protocols4, has required use of IPMVP methodologies by its evaluation, measurement, and verification (EM&V) providers. There are few RCx projects that follow-up after implementation to demonstrate actual savings using industry-standard M&V methods. These methods require measurements both before and after measure implementation, and adjustments to bring the pre- and post-implementation energy use to the same set of conditions. Under public-goods sponsored RCx programs, the savings verification function is often left to third party program evaluators, who traditionally have been allowed to propose their own savings verification methods. They must devise a methodology for a large population of buildings and measures, collect project information from program implementers, understand the building systems and how the measures affect them, somehow acquire baseline data for systems where measures have already been installed, and numerous other issues. This necessary and required process has many pitfalls that can be overcome with

1

Mills, E., H. Friedman, T. Powell, N. Bourassa, D. Claridge, T. Haasl, and M.A. Piette, 2004. “The CostEffectiveness of Commercial Buildings Commissioning: A Meta-Analysis of Energy and Non-Energy Impacts in Existing Buildings and New Construction in the United States.” LBNL-56637. 2 available at www.evo-world.org. 3 available at www.ashrae.org 4 available at www.calmac.org Jump et al: Tracking the Benefits of RCx.

3

National Conference on Building Commissioning: May 2-4, 2007

more thorough integration of M&V into RCx processes. This integration has additional benefits, as will be discussed later in this paper. Each of the involved parties in the implementation of RCx programs is motivated to ensure that savings are validated and will persist over time. The involved party and their motivations include: • • • • •

RCx service providers need an independent check on their energy savings estimates for quality assurance purposes, and need a means to demonstrate measure effectiveness. Building owners and their property management firms need to validate their investments in energy savings, and gain assurance that savings will persist over time. Utilities and third party RCx program implementers need assurance that their claimed RCx program savings stand up to third party review and that savings persist over their proposed lifetimes. EM&V contractors who are hired to evaluate RCx programs per CPUC requirements need a standard method of verifying the program implementer’s energy savings claims that is independent, accurate, repeatable, and defensible. Government agencies, such as the California Public Utilities Commission (CPUC), the California Energy Commission (CEC), and Independent System Operator (ISO), require verified and accurate savings, savings lifetimes, and program cost-effectiveness, to select effective programs and to enable accurate forecasting and planning activities to be carried out.

In order for RCx programs to be successful, a high level of confidence in their claimed energy savings and cost-effectiveness must be demonstrated. The UC/CSU/IOU’s monitoring-based commissioning (MBCx) program provided the venue to demonstrate the effectiveness of integrating M&V into RCx projects. Under this program, the addition of points and trending to the building’s energy management system (EMS) facilitated M&V and provided operators with tools to track the day-to-day energy performance of their systems.

Basic M&V Requirements The basic concept of quantifying energy savings under the IPMVP’s framework is summarized in its Chapter 3 opening paragraph: Energy or demand savings are determined by comparing measured energy use or demand before or after implementation of an energy savings program. In general: Energy Savings = Baseyear Energy Use – Post-Retrofit Energy Use ± Adjustments The “Adjustments” term in this general equation brings energy use in the two time periods to the same set of conditions. Figures 4 and 6 in this paper provide visual demonstrations of this concept. As described in our paper for NCBC 06, there is a great overlap in activities and data requirements for RCx and

Jump et al: Tracking the Benefits of RCx.

4

National Conference on Building Commissioning: May 2-4, 2007

M&V activities. Table 1 provides a list of the IPMVP’s requirements for M&V plans. Several M&V activities are part of RCx processes, and are documented in RCx plans. That there is a significant overlap demonstrates that integrating M&V into RCx is not an arduous task.

M&V Approach for UC Berkeley We completed two MBCx projects at UC Berkeley. The M&V plan was similar for each project, and was described in our NCBC 06 paper. In summary, it was based on the IPMVP’s Option B Retrofit Isolation method. The M&V Plan defined individual building systems and equipment (e.g. chilled water system, air handling system, etc.), and what data to collect and use for developing baseline energy use models. For each defined system, the plan described how models would be developed, what independent variable data would be used to make baseline energy use projections, how to assess uncertainty, and how the methodology may be used to monitor and maintain savings. UC Berkeley‘s campus-wide EMS has numerous monitoring points throughout its campus buildings. Each point is trended at 1-minute intervals and 6 months of data is stored and available for use. For the baseline period, we used data loggers and power measurement instruments to collect additional energy data and create “proxy” variables from equipment ON/OFF status and VFD speed feedback signals. Power and VFD speed points have been added to the EMS to facilitate energy calculations in the post-installation period. The M&V procedures made extensive use of the historical trend files, and established simple procedures for pulling data and performing the required analysis and reporting. The method may be programmed into the EMS or used in associated energy information software that accesses EMS data.

Jump et al: Tracking the Benefits of RCx.

5

National Conference on Building Commissioning: May 2-4, 2007

Table 1. M&V Plan Requirements for Adherence to IPMVP5 IPMVP M&V Plan Requirements Description of ECM and intended result Description of ECM boundaries Description of baseyear conditions and energy data energy and demand profiles occupancy type, density, and periods space conditions for each operating period equipment inventory equipment operating practices significant equipment problems Description of planned changes to baseyear conditions Identification of the post-retrofit period Establishment of conditions to which all energy measurements will be adjusted Documentation of the procedures used to verify proper installation of the ECM Specification of IPMVP Option Specification of data analysis procedures Specification of metering points and meter characteristics Option specific information (e.g. simulation software used under Option D) Specification of quality assurance procedures Quantification of expected accuracy associated with the measurement and analysis Specification of how results will be reported and documented Specification of data available for savings verification by a third party Descriptions of how non-routine baseline adjustments will be made Definition of budget and resource requirements for the M&V effort

RCx Plan x x x x x x x x x x

x

x

x

Summary of M&V Results Soda Hall Soda Hall houses UC Berkeley’s computer science department, and has five above-grade, and two below-grade floors totaling approximately 109,000 square feet. It provides faculty and staff offices, computer laboratories, several independently cooled computer and data server rooms, as well as an auditorium for lectures. The offices are located around the building’s perimeter, and the computer rooms are located in its core. Three main air-handling units (AHU) provide space conditioning in the building. A separate system serves the auditorium, and was not investigated. Two AHU serve the east and west side of the building’s perimeter zones, and one large AHU serves the core. The perimeter AHU are single-duct VAV systems with hot water coils in the VAV boxes. The AHU serving the core has a heating coil at the supply fan. Each AHU has chilled water coils at the main supply fans. In each AHU, supply fan VFD speed is modulated to maintain static duct pressure in the downstream supply ducts. To maintain building pressure, return fan speed was controlled to be 10 percent slower than supply fan speed. The chilled water is delivered to the coils through a constant speed primary and variable speed secondary chilled water loop. Two 215-ton water-cooled screw chillers (one primary, one backup), generate the chilled water. Two 2-speed cooling towers provide condenser water to the chillers and also provide cooling to the computer room AC units on several floors in the building 5

please see IPMVP, section 3.3.

Jump et al: Tracking the Benefits of RCx.

6

National Conference on Building Commissioning: May 2-4, 2007

core. During the investigation, it was discovered that the chilled water distribution extended to the neighboring building to provide cooling through fan-coil units in various classrooms, computer labs, and server closets. A steam-to-hot-water heat exchanger is used to generate hot water for the heating coils in the AHU and VAV boxes. Two variable speed pumps, one primary, one backup, circulate hot water throughout the building. Table 2 shows the list of systems defined for M&V purposes in Soda Hall, the available EMS points for each piece of equipment, and whether the defined system is affected by an installed RCx measure. The building’s electric and steam services are independently metered. The chiller’s electrical power is trended through the EMS, as is ON/OFF status of the constant speed equipment. Actual VFD speed signals from the supply and return fans were added to the system and trended. This data was used to determine energy consumption of all the targeted systems within the building. Several recommendations were made to improve energy performance in Soda Hall. The building was apparently designed to anticipate high internal loads from students and computer equipment. Minimum VAV box damper positions were set at 50% open, in order to maintain large supply air volumes. After the building was completed, the internal loads were not as high as anticipated, with students able to work at home via modems, and due to “cooler” computer technology. VFDs on the supply and return fans were not functioning properly, and were recommended to be replaced. Table 3 provides a summary of the measures that were implemented. The savings in this table were estimated using the eQUEST building simulation software6.

6

available at: www.energydesignresources.com

Jump et al: Tracking the Benefits of RCx.

7

National Conference on Building Commissioning: May 2-4, 2007

Table 2. Systems Defined for Soda Hall and Available EMS Points System

Equipment

Whole Building Main Electric Meters (2) Main Steam Meters (2) Chilled Water System Chillers 1 and 2 Primary Chilled Water Pumps P-5, P-6 Secondary Chilled Water Pumps, P-3, P-4 Condenser Water System Cooling Towers Condenser Water Pumps P-7, P-8 Air Distribution System AHU-1, SF-11, EF-12, EF-13 AHU-2, SF-14, EF-15 AHU-3, SF-16, SF-17 AHU-4, SF-18, SF-19 AHU-5, SF-20 Chiller Room Fans Chiller Room 181, SF-2, EF-2 Chiller Room 179, SF-3A, SF-3B, EF-1A, EF-1B

Affected by ECM? X

Available Points

kW lb X kW Status VFD speed X High/Low Status Status X VFD speed Status VFD speed VFD speed Status Status Status

AC Units Condenser Water Pumps P-9, P-10 AC-31 through AC-41 Hot Water System Hot Water Pumps P-1, P-2

VFD speed Status X VFD Speed

The bulk of the energy savings was expected to result from reduced fan energy use, with a limited amount of savings from the chilled water plant and steam heat exchanger. Implementation took place in the spring and fall of 2006. We determined the daily energy use of each defined system as the basis for modeling baseline energy use. Scatter plots of the daily energy use and daily average outside air temperatures for the whole building electric and steam, and peak period only electric, are shown in Figures 3a-c. Figure 3d shows the sum of energy use for combined affected systems, collectively labeled the HVAC system. These figures also show how the data was modeled, in order to be able to project the baseline into the post-installation period. It also shows models of the post-installation period data, for purposes to be described later in this paper. A simple linear regression between the daily energy use variable, and the average daily outside air temperature was the only model used for this effort. We intentionally kept the model as simple as possible, and relied only on data that is readily available through the EMS. There was a lot of scatter in the whole building electric and whole building steam plots, resulting in a poor fit using linear models. When we could not define a baseline model that met reasonable curve fitting criteria, we used an average of the daily values, such as that shown in the whole building steam plot. To resolve these issues, more data is required, and a re-examination of the model in terms of independent variables and equation form is required.

Jump et al: Tracking the Benefits of RCx.

8

National Conference on Building Commissioning: May 2-4, 2007

Table 3. Installed Measures at Soda Hall Estimated Savings Measure No.

Description

Estimated Measure Cost, $

Payback, yr

Implementation Date

Energy, kWh/yr

Energy, lbs/yr

Dollars, $/yr

266,250

$19,004

$1,550

0.1

$4,460

$7,000

1.6

AHU1-2

Resume supply air temperature reset control and return economizer to normal operation

10/25/2006

129,800

AHU1-3

Repair/replace VFDs in return fans

10/25/2006

34,308

AHU1-4

Reduce high minimum VAV box damper position

3/9/2006

46,300

119,300

$6,973

$15,250

2.2

AHU3-2 & AHU4-2

Option 2: Reduce high minimum VAV box damper position

3/9/2006

30,600

2,328,100

$22,603

$17,250

0.8

10/25/2006

242,000

$31,460

$14,000

0.4

AHU3-3 & AHU4-3

Re-establish scheduled fan operation and VAV AHU3 (includes repair/replace VFD on return fan EF-17), AHU-4 (includes repair/replace VFDs on supply SF18 and return EF-19 fans, and elimination of low VFD speed setting during the day)

$55,050

0.7

Total Percentage Savings Utility Data

483,008

2,713,650

$84,500

10%

51%

14%

Steam Electricity Cost

5,325,717 4,871,678 $621,575

Figure 3. Soda Hall Scatter Plots a. Total Building Electric

Jump et al: Tracking the Benefits of RCx.

b. Building Steam

9

lbs kWhr

National Conference on Building Commissioning: May 2-4, 2007

c. Peak Period Building Electric

d. HVAC System Electric

The HVAC system electric data had much less scatter than the whole building electric data. Whole building electric data includes lighting, plug, and other loads not affected by the RCx measures. In addition, there may be other effects not accounted for with a model based on outside air temperature, such as occupancy. We used the HVAC system model to verify savings. As can be seen in Figure 3d, it was possible to develop a more robust model of the baseline. Using these relationships, and typical mean year weather data for Oakland/Berkeley, we estimated baseyear and post-installation energy use, and determined energy savings by subtracting one from the other. Table 4 provides a comparison of the estimated and verified savings. Table 4. Comparison of Estimated and Verified Savings for Soda Hall Source kWh kW Lbs. Steam

Verified Savings**

Estimated Savings*

Whole Building

HVAC System

483,008 2,713,650

216,716 22 854,407

462,472 50

* based on eQUEST model ** based on baseline and post-installation measurements and TMY OAT data

The baseline model for the HVAC systems can be used as a tool to monitor the on-going performance of Soda Hall’s HVAC systems. Figure 4 provides a demonstration of the type of graphical information this tool would produce. The graphs show data in the post-installation period. The difference between the baseline model and the daily energy use bars shows the energy savings achieved each day. An error analysis on the savings would be based on the standard error of the regression, or RMSE value. The blue line is a model based on postinstallation data. When daily energy use consistently exceeds this line, savings are diminished, and operators can be alerted to remedy the problem. This simple graphical representation demonstrates how the basic M&V approach can effectively address the issues and requirements of RCx providers, customers, program implementers and evaluators, and regulatory agencies described in the introduction.

Jump et al: Tracking the Benefits of RCx.

10

National Conference on Building Commissioning: May 2-4, 2007

Figure 4. Baseline and Post-Installation Projections – Soda Hall HVAC Systems 4,500

4,000

Baseline Model 2 R =~0.50 RSME =~240 kWh

70

Post Model 2 R =~0.50 RSME =~260 kWh

60 3,500 50

40

2,500

Deg. F

Daily kWh

3,000

2,000

30

1,500 20 1,000 10 500

-

10 /3 1/ 2 11 00 /1 6 /2 11 00 /2 6 /2 11 00 /3 6 /2 11 00 /4 6 /2 11 00 /5 6 /2 11 00 /6 6 /2 11 00 /7 6 /2 11 00 /8 6 /2 11 00 /9 6 11 /20 /1 06 0 11 /20 /1 06 1 11 /20 /1 06 2 11 /20 /1 06 3/ 11 20 /1 06 4 11 /20 /1 06 5 11 /20 /1 06 6 11 /20 /1 06 7 11 /20 /1 06 8 11 /20 /1 06 9 11 /20 /2 06 0/ 11 20 /2 06 1 11 /20 /2 06 2 11 /20 /2 06 3 11 /20 /2 06 4 11 /20 /2 06 5 11 /20 /2 06 6 11 /20 /2 06 7 11 /20 /2 06 8/ 11 20 /2 06 9/ 20 06

0

Date HVAC Daily kWh Usage

Daily Oat Ave

Post Model

Base Model

Tan Hall Tan Hall houses UCB’s Chemistry and Chemical Engineering Departments, and has seven above-grade and two below-grade levels, totaling 106,000 square feet. It provides chemistry research laboratories, laboratory offices, classrooms, and offices for faculty and staff. This project was to be completed within three months, with recommended measures installed. In the basement, four 100-hp supply fans provide 100% fresh air to a large plenum. Two supply air risers deliver air from this plenum to each floor. Three sets of heating and cooling coils are located before the supply fans. Each fan has VFD that control fan speeds based on static pressure in the supply duct risers. Four VFD-controlled 60-hp exhaust fans are located on the roof. There are two exhaust duct risers and two exhaust fans associated with each riser. Air flow and distribution within laboratories and offices was not investigated in this project. A single 475-ton water-cooled centrifugal chiller provides chilled water to a constant volume primary loop. Two 20-hp constant speed chilled water pumps, one primary, one backup, maintain circulation in the loop. Two constant-speed condenser water pumps, one primary, one backup, deliver condenser water to the chiller from a cooling tower located in an adjacent building. The cooling tower was not included in the study. There were equipment cooling

Jump et al: Tracking the Benefits of RCx.

11

National Conference on Building Commissioning: May 2-4, 2007

systems and spot cooling systems for cold room refrigeration equipment and process labs in the building. These systems were not included in the investigation. As in Soda Hall, a steam to hot water heat exchanger is used to generate hot water, which is then circulated to heating coils in the perimeter zones throughout the building. The main AHU uses steam heating coils. Table 5 shows the list of systems defined for Tan Hall, and the available EMS points for each piece of equipment, and whether the defined system is affected by an installed RCx measure. Table 5. Systems Defined for Tan Hall and Available EMS Points System

Equipment

Whole Building Main 480/277 Electric Meter Main 220/110 Electric Meter Main Steam Meter Chilled Water System Chiller (VS) Primary Chilled Water Pumps CHWP-1, CHWP-2 (CS) Condenser Water System Condenser Water Pumps CDWP-1, CDWP-2 (CS) Cooling Tower (CS, 2-speed) AHU-3 AHU-3 Supply Fans SF-1, SF-2, SF-3, SF-4 (CS) AHU-3 Exhaust Fans EF-1, EF-2, EF-3, EF-4 (CS) Terminal Boxes and Fume Hoods associated with AHU-3 AHU-1 AHU-1 Chemical Storage AH-1, SE-1 AHU-2 AHU-2 Chemical Storage AH-2, SE-2 Heating Water System Heat Exchanger HWC-1 Hot Water Pumps HHWP-1, HHWP-2 Lighting System Lighting Circuits Plug Loads Plug Load Circuits Domestic Water Domestic Water Pumps

Affected by ECM? X

Available Points kW kW lbs/hr

X kW Status X Status Not Avail. X S/S & Speed S/S & Speed NA Status Status X Status NA NA NA

The building’s electrical and steam services are independently metered. The chiller’s electrical power is trended through the EMS, as is ON/OFF status of the constant speed equipment and actual VFD speed signals from the supply and return fans. Four main recommendations were made to improve operations and energy performance in Tan Hall. These are shown in Table 6. Both sets of chilled and condenser water pumps were found to be operating simultaneously, we recommended that one of the pumps be shut off and that the system be rebalanced by opening the balancing valve to achieve the required flow. The outside air lockout temperature function not functioning, and its set point was too high. The lockout Jump et al: Tracking the Benefits of RCx.

12

National Conference on Building Commissioning: May 2-4, 2007

temperature and the supply air temperature were both lowered by two degrees. Finally, a faulty steam valve controller was found that allowed steam to leak into the heating coil when it was not needed. This controller was replaced. The savings were determined using monitored building data and engineering analysis in spreadsheets. Table 6. Installed Measures at Tan Hall Estimated Savings Measure No. CHW-1 & CW-1 CH-5 CH-6 AHU3-4

Description

Eliminate simultaneous chilled water pump and condenser water pump operation Chiller outside air lockout temperature set point not operating Chiller outside air lockout temperature and Sat set point change Eliminate simultaneous heating and cooling Total Percentage Savings

Implementation Date

Energy, kWh/yr

Energy, lbs/yr

Peak Demand, kW

Dollars, $/yr

Estimated Payback, Measure yr Cost, $

5/24/2006

68,849

-

33.3

$9,639

$2,900

0.3

12/22/2005

361,184

-

0.5

$29,046

$0

-

6/26/2006 5/24/2006

119,767

-

8.3

$16,767

$0

-

103,775

10,543,991

48.8

$98,880

653,575

10,543,991

90.8

$154,332

$2,900

0.02

14%

62%

19%

Utility Data Steam

17,139,542

lbs

Electricity

4,720,647

kWhr

Cost

$798,007

The energy savings were expected to result from reduced chiller and pump operation, and reduced steam use. Implementation took place in the spring of 2006. Similarly to Soda Hall, we defined daily electric energy and steam use as the basis for the M&V analysis, and looked for relationships between whole building and system-level daily energy use, and the average daily outside air temperature. More robust linear relationships were found for whole building electric and steam, peak period electric, and chilled water system energy use, as shown by the scatter plots of Figures 5a-d. Using these relationships, and typical mean year weather data for Oakland/Berkeley, we estimated baseyear and post-installation energy use, and determined energy savings by subtracting one from the other. Table 7 provides a comparison of the estimated and verified savings. Table 7. Comparison of Estimated and Verified Savings for Tan Hall Source kWh kW Lbs. Steam

Verified Savings**

Estimated Savings*

Whole Building

HVAC System

653,575 91 10,543,991

663,184 69 5,995,232

686,519

* based on engineering calculations ** based on baseline and post-installation measurements and TMY OAT data

Jump et al: Tracking the Benefits of RCx.

13

National Conference on Building Commissioning: May 2-4, 2007

Figure 5. Tan Hall Scatter Plots a. Whole-Building Electric

b. Whole-Building Steam

c. Peak Period Electric

d. Chilled Water System Electric

The baseline and post-installation models for the whole-building electric, steam, and chilled water systems can be used to monitor the on-going performance of Tan Hall’s systems. Figures 6a-c demonstrate the graphical information this tool would produce. As in Soda Hall, this information can be used to track energy savings, levels of savings uncertainty, determine when savings are diminished, and identify the appropriate systems when problems arise.

Jump et al: Tracking the Benefits of RCx.

14

National Conference on Building Commissioning: May 2-4, 2007

Figure 6. Baseline and Post-Installation Projections a. Tan Hall Whole-Building Electric 25,000

80

Baseline Model R2 = 0.75 RMSE = 581 kWh

New Baseline R2 = 0.88 RMSE = 438 kWh

20,000

70

60

Daily kWh

40

10,000

Deg. F

50

15,000

30

20 5,000 10

0

0 6/1/2006

6/3/2006

6/5/2006

6/7/2006

6/9/2006

6/11/2006

6/13/2006

6/15/2006

6/17/2006

6/19/2006

6/21/2006

6/23/2006

6/25/2006

Date Total Building Daily kWh

Average Daily Outside Air Temperature

Baseline

New Baseline

b. Tan Hall Whole-Building Steam 90,000

80

Baseline Model 80,000

70

70,000

60

New Baseline 60,000

50,000 40 40,000 30 30,000 20

20,000

10

10,000

0

0 6/1/2006

6/3/2006

6/5/2006

6/7/2006

6/9/2006

6/11/2006

6/13/2006

6/15/2006

6/17/2006

6/19/2006

6/21/2006

6/23/2006

6/25/2006

Date Steam

Jump et al: Tracking the Benefits of RCx.

15

Baseline

New Baseline

Average Daily Outside Air Temperature

Deg. F

Daily Steam lbs

50

National Conference on Building Commissioning: May 2-4, 2007

c. Tan Hall Chilled Water System 80

6,000

Baseline Model R2 = 0.87 RMSE = 486 kWh

70

5,000 60

New Baseline R2 = 0.95 RMSE = 251 kWh

50

40

3,000

Deg. F

Daily kWh

4,000

30 2,000 20 1,000 10

0

0 6/1/2006

6/3/2006

6/5/2006

6/7/2006

6/9/2006

6/11/2006

6/13/2006

6/15/2006

6/17/2006

6/19/2006

6/21/2006

6/23/2006

6/25/2006

Date Chilled Water System

Average Daily Outside Air Temperature

Baseline

New Baseline

Discussion Overall project cost data from UC Berkeley are shown in Table 8. It shows costs for additional metering (mainly whole-building electric and steam, but some additional system-level points), agent costs, and in-house costs including those for measure implementation. Table 8. Project Costs Building Soda Hall Tan Hall

Metering Costs $ 4,442 $ 22,573

MBCx Agent Costs $ 62,160 $ 53,000

In-House Costs $ 51,087 $ 15,300

Total $ $

117,689 90,873

Several issues were addressed via the M&V procedures developed and implemented in the Soda and Tan Hall projects. These include: Focus resources on savings verification. In more traditional RCx programs, much time and effort is required to gather the data and estimate energy savings in order to determine which measures are cost-effective. This is often a criterion for owners to select measures for installation. However, measure costs are comparatively low, and it would be better to focus the limited project resources on metering and the verified savings approach. This has numerous additional benefits, such as more rigorous savings analysis, establishment of an energy tracking system, and provides diagnostic benefits.

Jump et al: Tracking the Benefits of RCx.

16

National Conference on Building Commissioning: May 2-4, 2007

Cost-effectiveness of MBCx approach. Including all the costs involved in the Soda Hall and Tan Hall projects, and based on verified savings; the projects remain very cost-effective with simple payback on investment of 1.7 years for Soda Hall, and 0.7 years for Tan Hall. The added costs of additional hardware for metering did not increase the project costs excessively. We also note that most large commercial buildings have the requisite whole-building electric interval meters as well as sophisticated energy management systems, so that applying this MBCx approach in the general building population should be economically viable as well. Savings estimations versus verified savings. Of interest to RCx providers and building owners investing in RCx projects, this M&V procedure provided an independent check of the estimated savings. Based on the data available, our estimations of electric savings were validated when compared to the verified savings. However, verified steam savings showed significant overestimation of our steam savings estimates. Based on this, the authors are reexamining the methods to improve savings estimations in the future. This also illuminates an issue program implementers have when submitting program results based on estimates. Even the best energy estimations rely on assumptions and methods that may ultimately result significantly different savings than what can be verified. Addressing savings persistence. The described M&V procedure lends itself well to programming into energy management systems or energy information software for helping operators monitor their building’s performance. Simple graphical representations such as those shown in Figures 4 and 6 enable rapid notification when energy performance in specific systems are degrading. Timely notification enables the problems to be mitigated before savings are lost. In addition, uncertainties in measurements and modeling can be tracked so that operators can know when problems are significant. An added benefit is that inclusion of additional monitored points and establishing an energy-tracking tool partially satisfies LEED-EB requirements.7 Smoothing the hand-off between implementers and evaluators. In California, with the requirement of evaluators to follow the IPMVP, EM&V teams are hard-pressed to collect the necessary baseline data in order to carry out this task. The method demonstrated here shows that M&V can be integrated into RCx projects, so that a proper M&V procedure can be carried out. If RCx providers are made responsible for carrying out IPMVP-adherent M&V procedures for their projects, EM&V teams can streamline their task by providing due diligence review of the individual M&V plans, data, and results, and use the more robust results to evaluate overall program effectiveness8. A significant barrier to integrating M&V into RCx projects exists. Savings verification is commonly viewed as a task to be completed at the end of a project, when in fact it must start from the beginning. There appears to be a limited pool of RCx providers, program implementers and evaluators who fully understand M&V procedures. Program implementers may need to include M&V training in their programs, and require RCx providers to develop and implement M&V plans as part of their projects.

7

In particular, the enhanced metering credits 5.1 – 5.3. See Leadership in Energy and Environmental Design for Existing Buildings (LEED-EB), U.S. Green Building Council (USGBC), www.usgbc.org. 8 This idea is echoed in Heinemeier, et. al. “Evaluating Retro-Commissioning Programs: The Challenges of Measuring Savings in a Measurement Based Program,” IEPEC Conference, New York, 2005. Jump et al: Tracking the Benefits of RCx.

17

National Conference on Building Commissioning: May 2-4, 2007

To understand these issues and barriers more fully, the California Commissioning Collaborative is sponsoring a project to evaluate existing methods and propose new methods for verifying savings within RCx projects. A similar project sponsored by NYSERDA is also underway9.

Conclusion The UC/CSU/IOU’s Monitoring-based Commissioning Partnership Program provided a welcome venue to demonstrate how readily M&V procedures may be integrated into RCx programs. RCx programs generally recommend operational changes that result in savings, while maintaining occupant comfort and other building services. Savings are based on estimates, and rarely verified. In the long run, this can lead to problems with the perception of RCx projects and programs. Monitoring-based commissioning programs provide the opportunity to develop tools to monitor and track savings, and notify operators when savings diminish. The two projects described here, with the added metering and analysis, remain cost-effective, and provide added benefits of rigorous savings verification, energy tracking, diagnostic capabilities, and long-term persistence tracking. This provides added security for owners, energy efficiency program implementers, and their regulatory agencies, that the savings are real and last over time.

9

California Commissioning Collaborative, www.cacx.org. and New York State Energy Research and Development Authority, www.nyserda.org.

Jump et al: Tracking the Benefits of RCx.

18