Accounting for customer solar energy
Looking for Insight
Utilities use Load Research to assign system and peak load responsibility across customer rate groups, or “customer classes.” These load research studies are very data-intensive, and provide inputs for rate design, cost-of-service, forecasting, procurement, distribution planning, and other utility functions. Most large utilities conduct them once a year, on the year following the study year. For example, most 2014 studies were conducted in 2015.
The estimates generated by these studies vary from utility to utility, but they have several common elements including:
- Quantifying customer class contributions at time of system peak (“in service territory X, medium commercial customers account for 40% of the annual system peak”),
- Determining customer class peak (“the residential class peaks at 6 PM”), and
- Providing the basis to estimate unaccounted-for energy (UFE).
By the same token, these studies utilize common inputs across all utilities including:
- Interval data sample (the basis of all load research),
- Billing data or other measures of annual sales – used to calibrate load profiles to known totals, and
- System load curves – the measure of the energy that was available on the grid.
For decades, these staples of load research remained fairly constant. In particular, interval data samples, which were expensive to install and to maintain, reflected the amount of energy that a customer drew from the utility. The cost of sampling has decreased dramatically with the advent of Advanced Metering Infrastructures (AMI) and with substantial increases in data processing capacity at reduced costs – but that is not the subject of this blog.
Paradigm shift in Load Research induced by Solar
The advent of solar energy is having a substantial impact in the way we conduct load research. It affects the inputs and the outputs of these studies in ways that require careful consideration that can change the meaning of the estimates. The interval data of solar customers no longer represents the load that is served by the utility, but also the energy that customers can put back into the grid. This creates a situation where a large group of the utility’s customer base increase the load that the utility can serve to other customers, but not in the same way that the system load is available to the utility – energy that is put back on the grid at the Distribution level cannot make it back to the Secondary or Transmission levels without upgrades to the infrastructure.
Starting from the beginning: a customer that is “net metered” (i.e. takes energy from the grid when the solar system produces less than the premise requires, and puts energy back on the grid when the system produces more than the premise requires) has two load components (power flows) that are utility-facing:
- Energy delivered by the utility (“Delivered”), and
- Energy received from the utility (“Received”).
Figure 1 illustrates these power flows.
“Net” is the subtraction of Received from Delivered, and it can be negative at a time when the solar system is producing energy and the premise is not using it. The flow of energy can be positive or negative several times during the same metered interval depending on what is going on at the premise (think, for example, of a hair dryer that turns on and off several times in 15 minutes). If only the Net energy is metered (the subtraction of all Received from all Delivered in the same interval), it will produce a true representation of the customer’s transactions with the grid, but not of what the utility ultimately delivered to the customer. The energy that the utility served the customer, regardless of what the utility took back from the customer, is represented by the Delivered energy flow. In a circuit that has several customers with solar generation, some of this Delivered electricity could be from other customers’ Received electricity.
Net versus Delivered – what is the right Estimate?
Utilities with large groups of solar customers have to determine what estimate best reflects the applications of their load research studies – one that uses “Net” energy as an input, or one that uses “Delivered” energy. Further, “Received” load profiles may be necessary, especially in areas of high Solar adoption, where “Received” could determine the size of the infrastructure needed to serve Solar customers. It may be possible that all are needed depending on the uses of the different stakeholders, which may create the initial impression that the utility has different estimates of the same load served to customers. These are actually estimates of three aspects of the same transaction: “Delivered”, “Received”, and “Net”.
This may sound trivial, but full visibility to all estimates simultaneously along with clear explanations of what each one is and what purpose it serves will go a long way in eliminating confusion regarding the availability and use of the different estimates. Some stakeholders will rely on their Load Research organizations to help them identify which is the best estimate for their purposes.
Importance of Load Research in the Solar World
The adoption of solar or other Distributed Energy Resources (DER) technologies will increase substantially over the next years. Load Research will play an important role to capture the impact of the new disruptive technologies for a variety of utility functions such as rate design, cost-of-service, forecasting, procurement and distribution planning.
The adoption of solar installations comes with a variety of changes that altogether drive the new look of load curves. More complex measurement configurations become necessary to record the different power flows from and to the premise. For example, Virtual Net Metering, a tariff arrangement by which California utility customers in multi-family buildings can benefit from solar generation that serves the whole building, can be configured as multiple Delivered-only accounts and one Received-only account. DNV GL can help utilities to successfully deal with the necessary adaptation to the new load behavior due to solar and other disruptive technology impacts. For more information about what DNV GL offers, read more here.