Small Commercial Needs Assessment. Part 1: Analysis of NSP Small Commercial Database
The Minneapolis Energy Office (MEO) has recently begun working with Northern States Power (NSP) to investigate the potential for a small commercial conservation program that would supplement NSP's current programs. NSP presently offers an informational mailing to non-demand customers and two types of audits for demand-billed customers, a free walk-through audit and a more extensive computer-assisted audit for which there is a charge based on the size of the audited facility. We envisioned a program that would supplement the free Quick-Check audit except that it would use different marketing methods, it would work with a somewhat broader range of customers (demand billed customers through 100 kW and any non-demand customers with high enough use to make the program cost effective), and it would add some combination of installation services, quality assurance inspections, and low interest financing or other financial incentives in order to increase the rate of actual installations achieved.
In order to determine the feasibility and optimum design of such a program, the MEO has undertaken a two part needs assessment. This report summarizes the first part, an analysis of the NSP commercial billing database for Minneapolis in order to estimate the size and composition of the probable target population, their amount and pattern of electric use, and what their electric use suggests in terms of operating patterns and end uses in each major type of business. The second part will be summarized in a separate report, and it concerns the results of 50 pilot audits and owner surveys undertaken to gather more data on ownership patterns, utility payment and metering patterns, and finances, as well as the end uses present and potential retrofit opportunities related to these end uses.
The Minneapolis commercial database provided by NSP includes some 14,534 cases from a total of over 16,000 in their commercial and industrial billing database. Cases were excluded that belonged to the industrial SIC groups (20 through 39), and cases with average demand over 100 kW were excluded unless they fell into the SIC groups most likely to contain non-profit organizations (80-86 and 91-93). Of the cases that met these screening criteria, 10,131 (69.7%) are non-demand billed and 4,403 (30.3%) are demand billed (see Table S.1 for details of these and all subsequent results in this summary). It was decided that an average electric bill of at least $800 per year would be required to make a program of this sort cost effective, which reduces the target population to about 6,385 cases, of which 2,094 (32.8%) are non-demand and 4,291 (67.2%) are demand-billed. This reversal of proportions comes from the fact that almost all demand-billed cases meet the target use criterion while only about 30% of non-demand cases do. The cases with adequate electric use data for our planned analyses comprised 4,945 customers, of which 1,646 (33.3%) were non-demand and 3,299 (66.7%) were demand-billed.
After some exploratory analyses it was decided to categorize businesses using a modified version of NSP's function code system, since it was more closely related to electric use than the federal SIC system. Some categories were collapsed together while a few others were split further using SIC code values as the criterion. Of the 15 categories that resulted, apartment buildings were already treated in MEO conservation programs and several others were too small or heterogeneous to be of much programmatic interest to us. This left 12 categories to be used for the analyses reported here, with a total population of 8,591 cases, or 4,900 within our target range of bill sizes and 3,811 that also have adequate electric consumption data. These numbers were considered to be large enough to merit a major program development and delivery effort and also large enough to permit the calculation of useful descriptive statistics both overall and by business classes.
Among the 12 business classes of interest, the largest groups are non¬food retail stores (1,710 cases out of the total population), office buildings (1,616), and business and social services (1,273). Somewhat smaller numbers of cases are found in restaurants and bars, personal services, multi-function buildings, warehouses, automotive services, churches and synagogues, and refrigerated food stores. Less than one hundred cases are found in the hotel and dormitory group and the hospital and nursing home group. In terms of average electric use the highest groups are the hospitals and nursing homes (12,543 kWh/month), refrigerated food stores (11,003 kWh/month), restaurants and bars (8,896 kWh/month), and warehouses (5,037 kWh/month), all of which are above the overall average use of 4,183 kWh/month. Based on aggregate use for each business class (thus taking both class size and average use into account) the most promising targets for marketing efforts appear to be restaurants and bars, office buildings, non-food retail stores, refrigerated food stores, business services, and warehouses.
We found a surprisingly high level of turnover among small businesses, with about 18% of all cases changing ownership in the last year alone. This fact has a number of important implications for the decision making of small entrepreneurs, the design of fiscally secure financing programs, and the potential for programmatic intervention at the time when businesses are being renovated or converted to a new use by new owners. Turnover varies considerably between business classes, being fastest in multi-function buildings, restaurants, office buildings, and non-food retail stores and being slowest by far in hospitals and nursing homes and in churches and synagogues.
Among demand billed cases, load factors (the degree to which average demand approaches the highest observed demand) increase with average use, ranging from 0.3 at 4,000 kWh/month (for a demand of 18 kW) to about 0.5 at 18,000 kWh/month (for a demand of 50 kW). This relation means that demand goes up with increasing use, but not as a constant proportion of use. Most business classes follow this overall relation closely and fall in a fairly modest range of values for the three measures, but there are several exceptions. Hospitals and nursing homes follow the same pattern but with unusually high values for all three variables. Churches (and to a lesser extent warehouses) have high demand relative to use, and thus low load factors, suggesting short hours of operation. Food stores have low demand relative to their use, and thus high load factors, which makes sense in terms of their constant refrigeration loads for food storage and their relatively long operating hours.
About one in four demand cases have demand that varies enough from month to month to generate ratchet charges. While a few of these cases have a substantial part of their demand charges due to ratcheting, in general ratcheting contributes only four to nine percent of their total billed demand and less than one percent of their total electric bills. Thus operational changes to reduce peak demand or to spread use more evenly will not be a high priority for small commercial cases overall, but they should be kept in mind for a minority of exceptional cases.
Small commercial customers show pronounced seasonality of electric use, with a short and steep peak in the summer that reaches levels well above the longer and more gradual winter peak. This suggests that the largest contributor to seasonality is the summer-peaking end use of air conditioning (and perhaps refrigeration to a lesser extent). Usually the potentially heavy winter-peaking loads such as space heating and water heating are handled with gas in the Minneapolis area, so the modest winter peak we observe comes from lighting seasonality and a very small admixture of cases with the larger winter uses. This overall pattern of strong summer peaking and lesser winter peaking holds true for a majority of business classes but there are some prominent exceptions. Churches and synagogues and auto services have lower summer seasonality and high winter seasonality, contributing to a pattern in which most business types show a negative correlation between the seasons (that is, if they have a strong winter peak their summer peak will be smaller, and vice versa). Warehouses occupy the middle range of this pattern, with modest seasonality both in summer and winter. Only the hotel and dormitory group deviates markedly from this pattern, since they have both high winter seasonality and the highest summer seasonality of any business class.
Full report (PDF)
Small Commercial Needs Assessment. Part 1: Analysis of NSP Small Commercial Database