This project focuses on designing and evaluating opportunities for point source-to- point source water quality pollution trading to meet a watershed-level defined total maximum daily load (TMDL) for phosphorus. We use the non-tidal Passaic River Watershed as a case study to investigate the magnitudes of the potential costs savings from a phosphorus trading program amongst the 22 major waste water treatment plants in the area.

Ambient water quality management at the watershed level is critically important to meeting the Clean Water Act (CWA) objectives and is an essential element of the on-going Total Maximum Daily Load (TMDL) approach which caps emission levels of both nonpoint and point sources of pollution. Cost-effective achievement of water quality goals, particularly through market-based approaches such as programs for pollution trading, has important impacts for the well-being of citizens throughout targeted watersheds, as well as for state and local budgets.

Utilizing load, price, and survey data for 119 large customers that paid competitively determined hourly electricity prices announced the previous day between 2000 and 2004, this study provides insight into the factors that determine the intensity of price response. Peak and off-peak electricity can be: perfect complements, substitutes, or substitutes where high peak prices cause temporary disconnection from the grid, as for some firms with on-site generation. The average elasticity of substitution is 0.11. Thirty percent of the customers use peak and off-peak electricity in fixed proportions. The 18% with elasticities greater than 0.10 provide 75% of the aggregate price response. In contrast to Industrial customers, Commercial/Retail and Government/Education customers are more price responsive on hot days and when the ratio of peak to off-peak prices is high. Price responsiveness is not substantially reduced when customers operate near peak usage. Diversity of customer circumstances and price response suggest dynamic pricing is suited for some, but not all customers. In the aggregate, we estimate that when peak to off-peak electricity prices reach a ratio of 5:1, the highest ratio during the study period, these customers would reduce their peak-period usage by about 10% reduction from their typical usage. These results are consistent with portfolio substitution elasticity estimates for large industrial customers on voluntary RTP tariffs at vertically integrated utilities, and suggest that the retail market context for RTP (i.e. two-part hedged vs. one-part market-based designs) does not greatly influence the price response behavior of industrial customers. In other words, The estimates of elasticities of substitution associated with two-part RTP service can be extended to similarly constructed RTP pricing plans in competitive markets where all load is exposed. These results add substantial support for the hypothesis that average price responsiveness conceals the diversity of response among customers even within the same business classifications, and that size along is not enough to predict price response.

We must know something about the specific customer characteristics or circumstances to understand why certain customers in a particular business class are price responsive and others are not. This is the type of information that is needed by energy suppliers in order to recruit customers that can provide sufficient load relief to promote market efficiency when prices are high or reduce the probability of an outage when there is a shortfall in system-wide reserves. Since the analysis of these data provide some of the only empirical evidence of the ability of large electricity customers to adjust electricity demand in response to competitive market prices, these results continue to be used by policy makers and analysts in price responsive rate design in electricity markets in California, New York, New England and elsewhere.