WolfeNotes.com

Martin: BPU/Rutgers EMP “one of the worst pieces of economic analyses I’ve ever seen”

Fresh off last week’s unprecedented and false attacks on DEP scientists and Senate Majority Leader Barbara Buono, DEP Commissioner Bob Martin repeats and expands his errors.

DEP Commissioner Bob Martin responds to critical questions before Senate Environment Committee during his confirmation hearing

Martin now is attacking the Board of Public Utilities staff, Rutgers economists and planners, and state econometric and energy models.

A “NJ Spotlight” story by former longtime Star Ledger energy and environment reporter Tom Johnson reported on the Assembly oversight hearings on how Governor Christie’s 90 day “reassessment” and more than $300 million cuts will impact the Energy Master Plan (EMP) – we wrote about the Assembly hearing here). In the Spotlight story, Martin blasted the economic analysis of the EMP.

But compare Martin’s hack attack with the professional response of his colleague, BPU President Lee Solomon (who merely put a happy face on a bad Christie policy):

Cabinet officials insist they do not envision a radical rewriting of the [Energy Master] plan. But Department of Environmental Protection Commissioner Bob Martin and Board of Public Utilities President Lee Solomon have made it clear they believe it fails to consider the economic consequences of pursuing such ambitious targets, including reducing energy consumption by 20 percent by 2020. …

“There has to be a cost benefit analysis on the things we do,” said Solomon, whose agency will take the lead in reviewing the plan. “We can’t simply impose what we would like to happen on the state of New Jersey.”

Martin is even more adamant about the plan’s flaws. “It is one of the worst pieces of economic analyses I’ve ever seen done,” he said at a clean energy summit in New Brunswick last month. “They didn’t put the numbers of what it would cost the ratepayer or industry.”

Like the bully on the playground, someone has got to take Mr. Martin on. He can not be allowed to go around trashing things and people he knows so little about.

Although I trained in planning at Cornell’s Graduate School and was a DEP planner and policy analyst for 13 years, I am no expert on the EMP and economic modeling. But it might as well be me because I don’t see any profiles in courage out there stepping up to the plate and taking on Bully Bob Martin.

The economic analysis of the EMP was conducted by Rutgers University (see: Updated Modeling Document) :

The Center for Energy, Economic and Environmental Policy (CEEEP) and the Rutgers Economic Advisory Service (R/ECON), both located within the Edward J. Bloustein School of Planning and Public Policy of Rutgers, the State University of New Jersey, have been tasked by the New Jersey Board of Public Utilities (BPU) to provide data and modeling support for the master plan effort.

The data for the U.S. used come from Global Insight, Inc., a national leader in economic forecasting

The economic analysis was based on econometric and energy models used widely in NJ. The modeling and economic analysis were extensively reviewed during a 2 year planning process:

A series of prior events helped to build the foundation for this report. On December 18, 2006, CEEEP and R/ECON presented the modeling framework used in this report to stakeholders. On January 5 and 19, 2007, CEEEP convened two technical working groups to elicit input on electric generation and transmission. In addition, CEEEP and R/ECON participated extensively in many stakeholder meetings convened as part of the Energy Master Plan process from late 2006 through September 2008.

The electric utilities and the business community participated in the EMP model development, planning, and economic analysis.

If the Rutgers economic analysis “was one of the worst pieces of analysis ever done” as Martin now claims, where were the energy industry experts and business community economists and why weren’t they raising objections to correct such a flawed piece of work?

A valid critique of the EMP analysis would focus on its failure to include billions of dollars in economic benefits and avoided costs of dirty coal power and global warming, which should be right up Martin’s alley as DEP Commissioner. But he is silent on these flaws because they make a stronger case for efficiency and renewables, while his objective is to gut those policies for short term economic rewards to the business community.

Martin is simply taking cheap shots by using after the fact economic conditions (i.e. dramatic drop in oil and gas prices; economic recession; reduced demand).

Martin’s severe criticism shows he’s not only a political cheap shot artist, but that he knows nothing about economic modeling, sensitivity analysis, scenario testing, or the role of models in planning. As the EMP itself explained, models are not precise and uncertainties are inherent in the modeling exercise:

In short, the Energy Master Plan must explicitly deal with uncertainty and the prospect that things will turn out differently from what was assumed. This often gets lost in the discussions as modeling is frequently assumed to be a forecasting effort with definite outcomes. The data and modeling assumptions have associated ranges of uncertainties. Even in situations in which one would think the range of uncertainty should be small, e.g., the cost of a combustion turbine, they can be surprisingly large. These uncertainties need to be considered when evaluating calculations. Although models calculate numbers to a precise value, this “precision” is a programming artifact and must be understood as such. What also should be kept in mind is that the range of uncertainty varies with specific assumptions. The uncertainty in the cost of a combustion turbine is smaller than the uncertainty of the cost of off-shore wind, which is in turn smaller than the uncertainty associated with the cost of a new nuclear power plant.

A primary driver for the current modeling draft calculations is the assumptions about the cost and magnitude of energy efficiency and demand response for electricity and natural gas. If one assumes that energy efficiency and demand response are cost-effective (which numerous studies have concluded) and that state policies can successfully influence energy efficiency and demand response, then one does not need modeling to conclude that energy bills will decrease, environmental impacts will be lessened, and the New Jersey economy will not be harmed. The modeling provides the order of magnitude, confirms the intuition, and helps target policies that can help to make these outcomes more likely. Thus, the preliminary calculations to date reflect the assumptions that they are based upon.

Here are the details, for Mr. Martin’s edification (and I question whether he has even read the EMP and reviewed the modeling):

C. Description of Models

Two major models are used as part of this effort. The first is R/ECON, a detailed econometric time series model of the New Jersey economy. The second is DAYZER, a sophisticated model of the PJM2 wholesale electricity power market. This model, DAYZER (Day-Ahead Locational Market Clearing Prices Analyzer), is a unit commitment and dispatch model that mimics, as closely as practical, the dayahead wholesale electricity market that New Jersey is part of (PJM), including calculating the locational marginal prices (LMPs) that vary by location and time. The results from DAYZER, along with many other assumptions, are then provided to R/ECONâ„¢ as inputs.

1. R/ECON – The New Jersey State Economic Model

R/ECON is an econometric model comprised of over 300 equations, which are solved simultaneously. The equations are based on historical data for New Jersey and the US. The historical data used to produce the model covers the period from 1970 to 2006, with some sectors updated through 2007. The sectors included in the model are:

ï‚· Employment and gross state product for 40 industries
ï‚· Wage rates and price deflators for major industries
ï‚· Consumer price index
ï‚· Personal income and its components
ï‚· Population, labor force and unemployment
ï‚· Housing permits, construction contracts, and housing prices and sales
ï‚· Energy prices and usage
ï‚· Motor vehicle registrations and stocks, and
ï‚· State tax revenues by type of tax, and current and capital expenditures.

The heart of the model is a set of equations modeling employment, wages, and prices by industry. In general, employment in an industry depends on demand for that industry’s output, and on the state’s wages and prices relative to the nation’s wages and prices. Demand can be represented by a variety of variables including (but not limited to) New Jersey personal income, NJ population, NJ sectoral output, or US employment in the sector. Growth in population is driven by total employment in the state and by state prices relative to national prices.

As part of this project the model was extended to include additional equations related to the energy sector. The equations in this new model sector are:

ï‚· Electric price per kilowatt hour, residential, commercial, industrial, and other;
ï‚· Electricity usage for residential, commercial, industrial, and other;
ï‚· Electric revenues in billions of dollars residential, commercial, industrial, and other;
ï‚· Natural gas price per thousand cubic feet, by sector, including the electric power sector;
ï‚· Natural gas usage by sector, including the electric power sector;
ï‚· Natural gas revenues in billions of dollars;
ï‚· Fuel oil price per million BTU, by sector;
ï‚· Fuel oil usage by sector;
ï‚· Motor fuel price and usage;
ï‚· Energy sales and corporate business taxes in millions of dollars; and
ï‚· Employment at electric utilities and other utilities.3

The R/ECON forecasting service produces four forecasts of the New Jersey economy each year. This study used the June 2008 R/ECONâ„¢ forecast as its baseline for the BAU and the June 2008 R/ECON Pessimistic forecast for the BAU Pessimistic baseline.4 Both baseline forecasts go out to 2020. The data for the U.S. used come from Global Insight, Inc., a national leader in economic forecasting. Tables 1 and 2 list the categories of inputs and outputs of the R/ECON model.

2. DAYZER – The PJM Wholesale Electricity Market Model

DAYZER calculates locational market clearing prices and the associated transmission congestion costs in competitive electricity markets.5 This tool simulates the operation of the PJM electricity market ”the dispatch procedures adopted and used by PJM”and replicates the calculations made by PJM in solving for the security-constrained, least-cost unit commitment and dispatch in the day-ahead markets. The LMP and congestion cost calculations are based on data on fuel prices, demand forecast, unit and transmission line outages, and emission permits costs. DAYZER incorporates all the security, reliability, economic, and engineering constraints on generation units and transmission system components.

DAYZER has the following features:

ï‚· Accurate security-constrained unit commitment and dispatch algorithms that mimics those used
by PJM in the day-ahead market
ï‚· Accurate modeling of PJM with its own particularities (second contingency constraints, locational
reserve markets, etc.)
ï‚· Captures marginal transmission losses in dispatch and clearing prices
ï‚· Captures transmission outages, transmission contingencies, nomograms, and planned and known
transmission upgrades
ï‚· Models accurately phase angle regulators and loop flows
ï‚· Allows users to analyze various scenarios and quantify the impact of key variables/assumptions
ï‚· Employs random outage using Bernoulli Probability modeling
ï‚· Enables the optimization of generation maintenance schedule based on reserves
ï‚· Uses import and export schedules to account for flows to and from neighboring markets

DAYZER requires that both transmission and generation additions and retirements be input exogenously into the model.6 The existing PJM transmission system is used in the DAYZER runs with additions as noted in Appendix A of this document.

In the current modeling effort, generation expansion plans are based on the following process: PJM’s load forecasts by zone by year are used to calculate the hourly loads using PJM’s 2006 load duration curve. The amount of system-wide installed capacity is calculated based on PJM’s 15% reserve margin. Renewable generation that is needed to meet individual states Renewable Portfolio Standards (RPS) is then included in the expansion plan. If additional generation is needed to meet the installed reserve margin, it is added. The type (baseload, intermediate, or peaking) and the fuel (nuclear, coal, or natural gas) are determined by reviewing the PJM generation interconnection queue for each particular PJM zone. Historically, the PJM generation queue contains more generation than is actually built. DAYZER is then run using the candidate expansion plan to ensure that generation unit capacity factors are appropriate for the type of unit and to ensure there are no hours in which demand exceeds supply in each zone that DAYZER tracks. In addition, locational marginal prices and net operating revenues are checked to ensure that either retirements or new generation would not otherwise occur. Modifications to the candidate expansion plan are made as necessary, and DAYZER is re-run until a satisfactory expansion is developed.