Last week the commissioners of the Texas Public Utility Commission once again complained about retail power suppliers who gamed the ranking system on the state’s retail electric power shopping website www.powertochoose.org. This post summarizes the problem and then offers a simple solution.
From the Houston Chronicle:
Texas’ utility commissioners complained Thursday about confusing or misleading rates many retail electricity companies offer consumers.
“The concept of choice doesn’t work if the customers aren’t educated about what they’re buying,” said Donna Nelson, chairwoman of the Public Utility Commission, which is investigating ways to make electricity shopping less onerous and confusing without greatly restricting the types of offers companies can make.
Commissioners want to improve the state’s Power to Choose website, which offers comparative pricing from more than 50 retail companies.
Nelson cited “deceptive” rate plans. She said retailers were warned about promoting misleading offers that promise low rates that are temporary or apply only if consumers use certain amounts of electricity.
“And they continue to do it,” she said, without naming specific companies.
As explained here last June (“Gaming the rankings on the Texas Power to Choose website“):
To help provide consumer information on competitive retail offers in the Texas electric power market, the Public Utility Commission of Texas maintains a website atwww.powertochoose.org. Enter your zip code, click a button, and it will display the top ten (out of nearly 300) offers.
Because the table shows the lowest priced offers first, with the average calculated assuming 1000 kWh of energy consumption, companies can compete for the front page by minimizing their average price at exactly 1000 kWh. But as it turns out, the low cost offer at exactly 1000 kWh of consumption may not be the low cost offer for consumers using a little bit less or a bit more than 1000 kWh.
When I searched for offers for the post last summer (June 2015) using a zipcode for downtown Houston, the top search result claimed a rate of 4.3 cents per kwh (whereas the state and national average price is around 12 cents per kwh). But that 4.3 cent rate was only obtained by a customer consuming between 999 and 1200 kwh per month, otherwise the customer would pay about 10.8 cents per kwh (lower than average, but much higher than it appears at first glance).
Just now (June 2016) I searched the Power to Choose website for downtown Houston, the top search results offers a rate of just 1 cent per kwh in large type! In slightly smaller type it indicates that the rate of 1 cent per kwh is at 1,000 kwh per month, while at 500 kwh per month the average rate is 11.7 cents per kwh and at 2000 kwh per month the rate averages 8.1 cents per kwh. That is quite a spread.
How do they do that? A click on the “Minimum usage fees / credits” link for the 1 cent offer reveals a $4.95 charge if fewer than 500 kwh are used, a $100 bill credit if usage is between 999 and 1501 kwh per month, and a $50 credit if usage is between 1500 and 2001 kwh. Clicking the “Fact Sheet” link for the offer reveals a bit more useful information. Fundamentally customers are charged 6.417 cents per kwh for energy use, and maybe they face the above described usage fee or credits, and they pay a base $1.50 per month charge and regulated TDU fees (passed through to customers at cost, in this case CenterPoint TDU fees are $5.47 per month plus 3.878 cents per kwh).
To assess how much the retailer is gaming the system, consider how much the rates vary for a level of consumption just one kwh different from the three points reported on the website: 500 kwh, 1000 kwh, and 2000 kwh.
Table 1: Calculated average rates at various levels of monthly consumption
| Monthly usage | 499 kwh | 500 kwh | 501 kwh |
| Average rate per kwh | 12.7¢ | 11.7¢ | 11.7¢ |
| Monthly usage | 999 kwh | 1000 kwh | 1001 kwh |
| Average rate per kwh | 11.0¢ | 1.0¢ | 1.0¢ |
| Monthly usage | 1999 kwh | 2000 kwh | 2001 kwh |
| Average rate per kwh | 8.1¢ | 8.1¢ | 10.6¢ |
Note: Calculations by the author based on the information described above.
While Donna Nelson opted not to name names, I’ll tell you the offer is the “PTC True Blue Infusion 3” plan from Infuse Energy, a 3 month, fixed rate offer with an 11 percent renewable content. Gexa Energy and Infinite Energy also offered rates that score at 1 cent per kwh, and a few more show up at 1.1 cent per kwh. These offers obtain their apparent low price by offering bill credits for consumption levels between 999 kwh and 2001 kwh.
What to do about gaming
Here is a straightforward approach to reducing the amount of gaming going on. Rather than calculate average rates as if the customer will consume exactly 1,000 kwh per month (or 500 kwh or 2,000 kwh), employ a seasonal consumption adjustment based on the month of the year. A Houston-based customer that averages a consumption level of 1,000 kwh per month over a year may use only 500 kwh of power during the mild spring and fall but 2,000 kwh or more in July and August. The state could calculate the average price for “1,000 kwh per month” using the estimated bills over the 12 month period assuming, as is reasonable, typical variation in consumption levels over the year.
Consider the estimated bills for twelve months of consumption which average exactly 1,000 mwh per month but vary in a normal fashion from month to month (assuming the rates, fees, and credits as described above).
Table 2: Estimated bills and average prices by month
| Month | Consumption | Estimated Bill | Average price |
| January | 600 | $68.74 | 11.5 ¢/kwh |
| February | 600 | $68.74 | 11.5 ¢/kwh |
| March | 500 | $58.45 | 11.7 ¢/kwh |
| April | 500 | $58.45 | 11.7 ¢/kwh |
| May | 900 | $99.73 | 11.1 ¢/kwh |
| June | 1,500 | $61.40 | 4.1 ¢/kwh |
| July | 1,800 | $142.28 | 7.9 ¢/kwh |
| August | 2,100 | $223.17 | 10.6 ¢/kwh |
| September | 1,700 | $181.99 | 10.7 ¢/kwh |
| October | 700 | $79.04 | 11.3 ¢/kwh |
| November | 500 | $58.45 | 11.7 ¢/kwh |
| December | 600 | $68.74 | 11.5 ¢/kwh |
| Totals | 12,000 | $1,169.04 | 9.7 ¢/kwh avg. |
Note: Calculations by the author based on the information described above.
Notice in no month was exactly 1,000 kwh consumed so the customer never achieved the 1 cent per kwh rate, even though it is possible. Note also that the average price over 12 months is better than the state average rate.
This “seasonal adjustment” approach would be slightly more complicated to deploy than the current system, mostly involving a simple bit of programming work up front and perhaps some continuing data updates, but the approach is objective, easy-to-understand, and has a clear payoff: consumers get better information.
This system, too, can be gamed by a clever retailer, but the difference between the advertised rate calculated using this method and the actual rates consumers face will be much smaller. With the payoff to gaming the rates smaller, maybe retailer marketing efforts get put to more productive use.