plant is actually operating near or at its design capacity. (cf Figure 11) Chronic operation at less than full utilization will result in higher construction and interest costs per unit treated than would be the case with a smaller plant. Second, similar considerations apply to the proposition that lower unit operating and maintenance costs will necessarily be achieved with larger plant sizes. From the discussion in a previous chapter, it is clear that lower unit operating and maintenance costs may not be achieved with a plant capacity in considerable excess of actual needs. In fact, it is generally the case that for any given actual flow that can be accommodated, operating and maintenance costs will be higher for a larger plant than for a smaller plant. Economies of scale in operation will be attained only if a treatment plant is operated near its intended capacity.

Finally, to build an overdesigned treatment plant in order to meet possible unexpected increases in demand is a one-sided strategy that ignores the full range of alternatives. The possibility that future demand might exceed forecasted demand arises because of the confidence with which the forecast is held. However, if a forecast is not held with certainty, then it is generally the case that future demand can fall short of the forecast with about the same probability as rising

above the forecast. What, then, are the alternative design strategies when demand forecasts are not held with certainty? On the one hand, a plant can be built to accommodate treatment needs in excess of currently forecasted needs. However, if actual future demand is not above forecasted demand, then the community incurs higher construction, operating, and interest costs on both a total and perunit basis than would be the case if a smaller plant had been built. On the other hand, a plant can be built to meet current and shortrange needs, say five to ten years, and the community can build increments to treatment plant capacity to meet additional needs as they occur. A potential loss is associated with this latter strategy, though; namely, if future demand is higher than forecasted, then the economies of scale associated with a larger plant have been foreUnder uncertainty, which of these two general strategies should be pursued? A recent study has indicated that the strategy of overbuilding treatment plant capacity in order to meet unexpected increases in future treatment needs is generally imprudent.* The rationale behind this finding is that, generally, the expected loss from building incrementally to meet short-term needs stemming from the potentially foregone economies of scale is less than the expected loss from overbuilding stemming from the potential higher costs of construction and operation.

Thus, economies of scale and safety margins are not, in and of themselves, sufficient economic justifications for overbuilding treatment plant capacity. Only if a community is expected to operate its treatment facility near full capacity within the near future, say five to seven years, will the potential cost savings be realized. In general, a strategy of building capacity to meet current and near-term needs will yield lower costs of construction and operation than the strategy of overbuilding.

PEAK LOADING

A community's hydraulic characteristics must be incorporated into the design characteristics of its treatment plant in order to attain

[p. 117]

target degrees of treatment. The expected peak load is one of the most important characteristics that must be considered in meeting design efficiencies of a plant on a continual basis. Peak loads can be met by a combination of three basic methods: varying detention times and recirculation rates, use of flow equalization devices or tanks to smooth the flow of influent and permit processing at non-peak

Ibid. pp. 1195-1206.

periods, and building sufficient operating capacity to handle peak loads as they occur.

If it is the case that anticipated peak loads are met primarily by building sufficient capacity to meet them as they occur, then this practice will contribute to the prevalence of stated excess capacity. To illustrate, suppose that two communities plan to treat the same average daily flow, say one million gallons per day, but that the first community has an average peak at a daily rate of 1.2 million gallons and the second has a peak of 2.0 million gallons. If these peaks are met solely by building capacity to handle them, then the first community will build a plant with a smaller design capacity than will the second community. Consequently, the first community's plant will have a higher calculated utilization rate (actual flow/design flow) than the second community's plant. From this example it can be seen that if it is common design practice to build enough treatment plant capacity to meet peak loads as they occur, then it might be expected that observed lower rates of utilization are associated with higher peak loads.

The validity of this partial explanation for the prevalence of excess capacity can be statistically tested by computing the correlation between the rate of capacity utilization and a measure of peak loading. A negative correlation between these two variables is expected if the practice of using excess capacity in order to meet peak loads is prevalent. The rate of utilization is measured by the ratio of actual average daily flow to design capacity and peak loading is measured by the ratio of peak load to average daily flow.

The statistical results are reported in Table 31. As can be seen by inspection of the first row of this table, the correlation between peak loading and utilization rates is negative but low (a value of -1.0 denotes perfect negative correlation, 0 is perfect non-correlation, and 1.0 is perfect positive correlation). Each correlation is, however, significantly negative (i.e., significantly below zero) by the usual tests of statistical significance. In the second row of the table the

TABLE 31.-STATISTICAL RELATIONSHIPS BETWEEN CAPACITY UTILIZATION
AND THE RATIO OF PEAK LOAD TO AVERAGE DAILY FLOW

[p. 118]

percentages of variation between plants in capacity utilization attributable to variation between plants in peak loadings are reported. The percentage of explained variation ranges from a low of 3.5 percent for stabilization ponds to a maximum of only 8.6 percent for the activated sludge process. In other words, less than nine percent of variation in utilization rate can be accounted for by peak loading, and so justifiable on an engineering basis. The remaining 90-odd percent is attributable to factors other than peak loading.

[p. 120] APPENDIX A-SURVEY QUESTIONNAIRE STUDY OF WATER POLLUTION ABATEMENT COSTS

GENERAL DIRECTIONS

A separate report should be prepared for each plant. It is necessary to know these data for each plant so as to relate the production and financial data to the wastewater abatement cost data when making cost burden and incentive analyses.

A plant is defined as the total facilities and operations at one location. Whether a few or many products are made at this location, it still should be considered one plant. This excludes facilities restricted entirely to such operations as warehousing and storage, research and development, and sales offices.

In the preparation of this survey questionnaire, care was taken to request information, wherever possible, in terms identical to those utilized in various reports to the Bureau of the Census. This was done to provide a recognized standard for some of the information requested and to permit the respondent to provide information similar to that which has been compiled for other reports. Please report for calendar year 1969 unless otherwise specified. If this is not possible, specify the reporting period for which data are provided.

Please return the completed form to Leonard Lund, National Industrial Conference Board, 845 Third Avenue, New York, New York 10022. Do not indicate your name or company on this form. The Code Number on this page identifies you to The Conference Board. No personal or corporate identification will appear in any report based on this survey without your explicit authorization.

ITEM 1. PRODUCT INFORMATION

(a) Principal product (s) of this plant

[p. 121]

(Describe by using categories defined in the Standard Industrial Classification Manual, e.g., "Chemicals and Allied Products," "Industrial Gases," "Food and Kindred Products, Fluid Milk," "Transportation Equipment, Motor Vehicles," or similar descriptive phrases.)

(b) Standard Industrial Classification Code (s). (If known)

(4 digit code(s) )