Low-Risk and Cost-Effective Prior Savings Estimates for Large-Scale Energy Conservation Projects in Housing: Learning from the Fort Polk GHP Project
- ORNL
- Thermal Energy Systems Specialists, Inc.
Many opportunities exist for large-scale energy conservation projects in housing: military housing, federally-subsidized low-income housing, and planned communities (condominiums, townhomes, senior centers) to name a few. Energy savings performance contracting (ESPC) is now receiving greater attention, as a means to implement such projects. This paper proposes an improved method for prior (to construction) savings estimates for these projects. More accurate prior estimates reduce project risk, decrease financing costs, and help avoid post-construction legal disputes over performance contract baseline adjustments. The proposed approach to prior estimates is verified against data from Fort Polk, LA. In the course of evaluating the ESPC at Fort Polk, Louisiana, we have collected energy use data - both at the electrical feeder level and at the level of individual residences - which allowed us to develop calibrated engineering models which accurately predict pre-retrofit energy consumption. We believe that such calibrated models could be used to provide much more accurate estimates of energy savings in retrofit projects, particularly in cases where the energy consumption of large populations of housing can be captured on one or a few meters. The improved savings estimating approach described here is based on an engineering model calibrated to field-collected data from the pre-retrofit period. A dynamic model of pre-retrofit energy use was developed for all housing and non-housing loads on a complete electrical feeder at Fort Polk. The feeder serves 46 buildings containing a total of 200 individual apartments. Of the 46 buildings, there are three unique types, and among these types the only difference is compass orientation. The model included the heat transfer characteristics of the buildings, the pre-retrofit air source heat pump, a hot water consumption model and a profile for electrical use by lights and other appliances. Energy consumption for all 200 apartments was totaled, and by adjusting thermostat setpoints and outdoor air infiltration parameters, the models were matched to field-collected energy consumption data for the entire feeder. The energy conservation measures were then implemented in the calibrated model: the air source heat pumps were replaced by geothermal heat pumps (GHPs) with desuperheaters; hot water loads were reduced to account for the low-flow shower heads; and lighting loads were reduced to account for fixture delamping and replacement with compact fluorescent lights (CFLs). Our analysis of pre- and post-retrofit data (Shonder and Hughes, 1997) indicates that the retrofits have saved 30.3% of pre-retrofit electrical energy consumption on the feeder modeled in this paper. Using the method outlined, we have been able to predict this savings within 0.1% of its measured value, using only pre-construction energy consumption data, and data from one pilot test site. It is well-known that predictions of savings from energy conservation programs are often optimistic, especially in the case of residential retrofits. Fels and keating (1993) cite several examples of programs which achieved as little as 20% of the predicted energy savings. Factors which influence the sometimes large discrepancies between actual and predicted savings include changes in occupancy, take-back effects (in which more efficient system operation leads occupants to choose higher levels of comfort), and changes in base energy use (e.g. through purchase of additional appliances such as washing machines and clothes dryers). An even larger factor, perhaps, is the inaccuracy inherent in the engineering models (BLAST, DOE-2, etc.) commonly used to estimate building energy consumption, if these models are not first calibrated to site-monitored data. For example, prior estimates of base-wide savings from the Fort Polk ESPC were on the order of 40% of pre-retrofit electrical use; our analysis has shown the true savings for the entire project (which includes 16 separate electrical feeders) to be about 32%. It should be noted that the retrofits carried out at Fort Polk are unlike most utility programs in that all of the family residences at the site received the package of retrofits. Thus in this study we were not concerned with issues of free riders or free drivers. However, like most military family housing - and quite unlike the housing in most utility programs - the family housing at Fort Polk is not individually metered. The evaluation program upon which the present work is based relied on data collected from monitoring equipment installed at the site.
- Research Organization:
- Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
- DOE Contract Number:
- DE-AC05-00OR22725
- OSTI ID:
- 978707
- Resource Relation:
- Conference: National Energy Program Evaluation Conference, Chicago, IL, USA, 19970801, 19970801
- Country of Publication:
- United States
- Language:
- English
Similar Records
The evaluation of a 4000-home geothermal heat pump retrofit at Fort Polk, Louisiana: Final Report
Geothermal heat pump energy savings performance contract at Fort Polk, LA: Lessons learned