Margaret F. Fels

The applicability of PRISM to electric heating and cooling

Abstract To explore the applicability of the scorekeeping method to electrically heated houses, variations of the basic PRISM model are tested on a New Jersey utility data base. Samples of houses with and without electric cooling are selected. The results indicate that the heating model applied to electrically heated houses without cooling performs as well as it has previously for gas-heated and oil-heated houses. The heating-plus-cooling model works well on houses with relatively strong cooling, while houses with erratic or weak cooling will require an alternative approach. The average cooling reference temperature estimated by the model is found to be well above the average heating reference temperature. Overall, the NAC index as the basis for savings estimates appears highly reliable of electrically heated houses.

Refraction of PRISM results into components of saved energy

Abstract A simple methodology is developed to augment the physical interpretation of PRISM results. With this “refraction” method, the estimated energy savings are decomposed into physically meaningful components, indoor temperature changes are inferred, and the effect of such changes on energy savings is quantified. The method is demonstrated in two test applications. The first, to individual-house billing data for a set of 243 gas-heated houses, suggests the dominant role of structural retrofitting in the energy savings achieved by weatherization of the houses. The second application, to utility aggregate data for nearly one million gas-heated houses in one state, points to the importance of lower thermostat settings in the years immediately following the oil embargo, and, since 1980, to a substantial decline in consumption by appliances including water heaters. More recently, the role of structural retrofitting seems to be increasing, but a “takeback” effect due to a concomitant increase in indoor temperatures may be eroding the resulting savings. While the refraction results must be interpreted with care, their physical reasonableness, in both test cases, reinforces the potential of energy billing data to enhance the understanding of energy conservation trends.

COMPARATIVE ENERGY COSTS OF URBAN TRANSPORTATION SYSTEMS

Abstract Several urban transportation systems are compared on the basis of the energy resources consumed in the manufacture of their vehicle and guideway systems and in their operation. Four systems, the auto, bus, rapid rail and personal rapid transit, are extensively analysed, while the energy characteristics of dial-a-ride systems and the motorcycle are estimated. Analyses of bicycling and walking are also included for a comparison of motorized modes with human propulsion systems. For each system the energy required to manufacture a vehicle, the energy required to manufacture the guideway, and energy of operation are estimated. The manufacture contributions are amortized over component lifetimes and added to the energy of operation to give an estimate of total energy consumed per vehicle-mile by a system. To provide a measure of the potential energy efficiency of the systems, the average energy consumed per available seat-mile is calculated to compare the systems when they are operating at capacity. An example containing assumed average occupancy levels gives a more realistic comparison on the basis of energy consumed per passenger-mile.

Using the scorekeeping approach to monitor aggregate energy conservation

Abstract A simple model for monitoring natural gas consumption in large aggregates of houses is developed by extending the individual-house scorekeeping approach. The model is tested on aggregate utility billing data for gas-heated houses in New Jersey. The resulting aggregate NAC index, representing pre-customer gas consumption in average weather, is extremely well determined, and the corresponding R2 values of the fits of the data are very high. From the model applied to sixteen years of monthly data, interesting conservation trends emerge. Since the oil embargo, a drop of about 30% in per-customer gas consumption in the State is evident; the most dramatic rate of decrease occurred in the first few years following the embargo. The individual PRISM parameters, though more erratic than the NAC index, suggest interesting structure in the year-to-year consumption changes.

Seasonality of non-heating consumption and its effect on PRISM results

Abstract By subtracting furnace submeter readings from whole-house meter readings, the variability of the energy consumed by appliances using the same fuel as the furnace is examined. Comparable analyses are performed on two sets of houses, one with gas and the other with electric heating. Although erratic month-to-month fluctuations occur in individual houses, median results for non-heating consumption show a strong sinusoidal pattern, for both fuels, with a peak in winter and a valley in summer. The effect of this seasonality on PRISM results is investigated. It is found that the models estimate of base-level consumption represents the level of consumption in summer while, on average, the increased non-heating consumption during the remainder of the year loads systematically onto the models estimate of heating consumption, without deteriorating its performance. A new method for approximating the resulting hidden non-heating component of consumption from whole-house billing data is proposed. Thus a new interpretation of Normalized Annual Consumption (NAC) is derived in terms of three components: base level, hidden non-heating, and heating consumption.

Measuring household fuel consumption on the Standard Living Cycle

A simple model for monitoring natural gas consumption in large aggregates of houses is developed by extending single-house analyses. The parameters of the model are baseload consumption, rate of fuel use per degree-day for heating, and a reference temperature. Accuracy of the results is assessed by means of confidence intervals derived for the nonlinear model.

Using billing and weather data to separate thermostat from insulation effects

A simple methodology, which uses aggregate utility sales data to measure and decompose residential conservation trends, is demonstrated. The data base is the aggregate natural gas consumption of the nearly one million houses in New Jersey which use gas for heating. The conservation index resulting from the analysis is normalized annual consumption, which adjusts per-household consumption to typical weather conditions. To smooth erratic year-to-year changes, four-year composite analyses are moved forward in time over the period 1970–1982.

Indices of energy consumption: An exploratory analysis of a utility's monthly billing data

Simple weather-sensitive models are derived and applied to monthly energy billing data for a large utility region in New Jersey. Natural gas demand in the residential sector is explored in depth, and the method is applied to other gas and electricity sectors. The results are related to an earlier empirical study of single-family residences.

Breakdown of rapid rail energy costs: A study of three systems

The total energy consumption of three urban rapid rail systems is analyzed to obtain a detailed breakdown of the traction and station contributions. The systems chosen, PATH, PATCO and BART, span a wide range of characteristics representing age, service region and size. The primary data source for the analysis is monthly bills, whose disaggregation determines the ultimate breakdown of the total energy consumption for each system.

The stability of PRISM estimates

Abstract The stability of the PRISM parameters against changes in time periods is studied, for a better understanding of the problems likely to emerge when meter readings are missing or when only bimonthly data are available. Consecutive monthly readings spanning intervals from 7 to 24 months, and one-year data sets with gaps, are examined. The results indicate that 12 monthly readings are optimal for the most reliable PRISM results. The NAC index is far less sensitive to missing or insufficient data than are the individual parameters, whose stability is affected by the seasons included or omitted in the estimation period. Results from less than ten consecutive readings, or from less than nine readings spanning a full year but with large gaps, should be scrutinized carefully. Reliable savings estimates are available from bimonthly data, as long as missing readings are not prevalent.