Looking in the Mirror at a Low Load Factor Customer

I don’t know if it was me looking back at my utility load forecasting and generation planning days or thoughts about proper sizing of microgrids work I had done recently, but I recently developed a desire to investigate my own electricity usage. Before I jumped in with both feet, I did a little investigating around typical residential usage as measured by average load factor. Load factor is defined as the average load divided by the peak load in a specified time period, commonly used as a measure of the utilization rate, or efficiency of energy usage. According to the U.S. Department of Energy, the average load factor of a residential customer in Phoenix, AZ is approximately 33-34%. Therefore, the average customer’s peak usage in my surrounding area is a little over three times their average daily usage. I was pretty sure (hoping) that my own usage pattern would be somewhat higher.

Fortunately, my house in Arizona has a smart meter with access to my hourly usage. So I set off to prove myself a conscientious user of electrons. It turns out that my load factor was not higher but was, in fact, LOWER than I would have expected. In fact, my load factor was just over 16%, or put another way, my peak usage was over 6 times my average usage.

After digging a bit deeper, I discovered that on average, my highest KWh load periods occurred only during a very small percentage of my overall usage duration. This is seen in the steep curve at the top left of the graph. In contrast, a perfectly balanced load factor customer’s usage would chart linearly. Since utilities must plan for peak load periods, this low load factor ratio impacts grid infrastructure sizing, and therefore the total cost of providing service — leading to higher rates. The unknown from my perspective is whether or not other residential customers behave similarly.

If my peak usage hours coincide with other residential customers in my area, then system planning at the local level is affected. So I was curious to see if my hourly energy usage behavior is abnormal. As I am not on a time-of-use (TOU) rate, I asked a friend who is if I could compare their electricity usage for the same period. To my relief, his load factor was also slightly over 16%.

Still, I wanted an explanation, so I put together a load comparison chart.

The TOU period from 2-8pm shows that my friend responded to the price change and decreased his usage to save. Since I am not on the TOU rate plan, I don’t have that incentive. Interestingly, however, we both experience peak usage after the TOU time period ended at 9pm. In my case, I lower my thermostat setting without thinking about the implications, and he likewise has postponed high-usage items due to higher TOU rates earlier in the day. Ironically, because I am not enrolled in the TOU rate program, I have higher usage during the TOU period.

So what does all of this mean? From a distribution planning perspective, assuming other residential customers have similar load profiles, my utility has to plan for a system sizing somewhere around 6 times average usage. At a minimum, this impacts the local electrical system sizing, though I don’t think it’s a big deal for generation planning since last year’s peak at my utility occurred between 5-6 pm. I should have stopped there, but I was curious, so I looked at CAISO system summer load data and found that the California peak coincided with mine. So, I guess if I had the same usage pattern in California, I would be impacting both generation planning and local distribution planning. (Who knew?)

So, what has my self-examination uncovered? My load profile is quite similar to my friend, even though we’re on two different rate plans. Our low load factors may be contributing to incremental distribution costs, but it’s unclear to what extent. Also, we’re probably not triggering additional peak generation assets due to being out of sequence with the current utility peak load period, but we may be adding generation costs.  So what now?

I think I’ll get an Electric Vehicle, see Part 2.