Projecting Forward Utility Costs

Most marketers of solar, energy efficiency and energy supply share one thing in common:   They like to predict that energy prices are destined to go up forever.

Sadly few people are knowledgeable about historical utility costs to push back against this nonsense.  See Myth Busters:  Utility Prices Always Go Up.   Even regulators who should know better allow aggressive marketers to woo customers with scary stories of endless price increases – when their own data shows that the highest utility prices  in the country, in the Northeastern US states, have moved up an average of less than 3% per annum for the past 20 years.

Whatever the past, however, decision-makers must make reasonable assumptions about future price increases.   Their projections will be instrumental in determining whether it is prudent for them to make investments in renewable energy projects such as solar or energy efficiency projects such as cogeneration, upgraded boilers, LEDs, or other energy conservation projects.

Accordingly, we must attempt to forecast future utility costs.  For this purpose, let us leave our crystal ball to one side and rely on statistical measures only.

Commodity or Generation Charge.    The cost of electricity is a function of fuel costs.  In the Northeast, natural gas drives electricity pricing; a generation of environmental regulation has shifted generation fuels from coal and fuel oil, which are high in pollution-causing sulfur dioxide, to natural gas and nuclear energy.  With some nuclear power plants retired (Vermont Yankee) or scheduled for close (Indian Point), natural gas is becoming more important as a driver of incremental electricity costs.

A ten year history of major New York utilities reflects a wide-range of generation charges.  These prices roughly follow the prevailing trends in natural gas pricing shown in the chart above.  Note in particular the peak in nat gas pricing in 2008 which drove generation charges to an all-time high.    Also note the contrast in generation charge among utilities themselves.

  National Grid Con Ed NYSEG
2004 0.0566 0.1014 0.0674
2005 0.0784 0.1396 0.0798
2006 0.0638 0.1152 0.0743
2007 0.0716 0.1154 0.0826
2008 0.0814 0.1280 0.0818
2009 0.0468 0.0852 0.0508
2010 0.0520 0.0930 0.0515
2011 0.0477 0.0820 0.0520
2012 0.0391 0.0701 0.0439
2013 0.0448 0.0994 0.0610


If we throw out the high and low prices and average the balance we have the following:

National Grid Con Ed NYSEG
0.0577 0.1039 0.0648


While the volatility of daily electricity pricing exceeds 100%, volatility – i.e., one standard deviation of price movements over time — is “smoothed out” when prices are set over long intervals, in the case of this data over a year.  Consumers in 2014, for example, experienced price spikes when utilities raised prices to pass through unexpected supply costs during the Polar Vortex of January and February.  The Vortex drove nat gas prices to close to $100/mmbtu, over ten times historic winter prices.  Electricity supply costs for generation increased from some $.10/kWh to over $.30/kWh.  Over the course of the year the impact of those sharp supply cost increases was muted.

In view of this “smoothing” effect, we propose a reduced rate of volatility of 30%.  At this rate, there is about a 2/3 chance that prices will stay within the following range – and a 1/3 chance that they will move outside it:

  National Grid Con Ed NYSEG
Low 0.040 0.072 0.045
High 0.075 0.135 0.084


Distribution costs.  If we take the same approach to distribution costs, we calculate the following ten year average for the three major New York utilities:

National Grid Con Ed NYSEG
0.037 0.075 0.024


Unlike commodity pricing, these distribution costs can be expected to increase over time by the rate of inflation.  After 10 years these costs will rise at a compounded rate of 5% to the following:

National Grid Con Ed NYSEG
0.060 0.122 0.039


Demand.   Demand charges for commercial entities are assessed on a monthly basis, not per kWh, and are calculated based on the maximum demand in any 15-minute period during the month.  If a commercial facility turns on its air conditioners in June, for example, its energy load will surge with a burst of energy required to start the motors.  The utility charges the consumer an amount that reflects the capital and maintenance costs necessary to maintain enough capacity to meet this surge in demand.

Utilities generally do not charge small customers for demand.  These customers do not operate heavy equipment that can put a sudden strain on the system.  Large commercial users, however – including large governmental facilities, non-profit institutions like schools, or large industrial plants – may put sudden load demands on the utility grid.  Since electricity cannot be stored, utilities must maintain sufficient capacity to meet these sudden shifts of load by the large customers.

The cost of demand varies.  In New York’s National Grid territory, for example, demand generally costs about $8 per kW.  Thus, if an energy user were to turn on its lights and hvac system and its simultaneous usage over a 15-minute period exceeded, say, 15 kWh, the demand charge for the month would be $120.   If the customer used an average of 10 kWh per working hour for the month for a total of 1,720 kWh and another 5 kWh for non-working hours for a total of 2,752 kWh, the demand charge per kWh for the month would be $.027.

Demand charges generally increase with inflation.   At the same time, we believe that energy efficiency programs – in particular the introduction of battery technology – can be expected over the next few years to reduce demand load.  Accordingly, in our analysis we assume that demand charges will remain constant and flat.

Note that our estimate of $.027 is based on very conservative assumptions.  Our example assumes a ratio of about 3:1 demand vs. average hourly consumption. Most commercial customers experience a much larger ratio of closer to 5:1.  In these cases, monthly demand charges divided by actual consumption will likely exceed $.03-.04/kWh.

Putting it all together.    Let us now combine our projected generation charges, distribution and demand costs for year 10.  For purposes of our analysis we will assume that commodity or generation increases at a rate of 2.0% per annum – less than the historical average over the past 20 years of 2.2% per year.  The anticipated range of prices for the largest New York utilities will be as follows:

National Grid Con Ed NYSEG
0.133 0.231 0.117
0.173 0.303 0.162


If we take an average of the range of anticipated ranges we have the following:

National Grid Con Ed NYSEG
0.153 0.267 0.140



We are often asked about our confidence level in these projections.  We answer that our confidence is similar to the probability that is assigned by statisticians to one standard deviation:  about 66%.  We believe that our assumptions are conservative and based on historical price movements.  Furthermore, we always remind our clients that prices are close to historic lows and accordingly that there is more upside price risk than downside opportunity cost.   At these low prices a change in the assumptions does not dramatically undermine the projected benefits.  The story would be different if prices were higher today.

At the same time consumers must be careful not to delude themselves and fellow stakeholders with rosy projections of guaranteed savings.  Increases in utility prices of 2.5-3% are likely but by no means certain.  A number of developments – e.g., sudden economic weakness, environmental disaster, or remarkable new technologies – could alter the economics of energy supply and demand quickly.

While most planners do not include outlier events in their scenarios consumers of commodities like electricity must always acknowledge these risks and speak about them openly.  Only then can they avoid the disappointment and political consequences that can come from thwarted expectations.