Utility processes and systems impacted by distributed PV
In previous blog posts, we’ve looked at how the growth of distributed PV can impact grid operation at the system level. These blogs addressed how distributed PV can lead to the often referenced “duck curve,” and how SolarAnywhere® FleetView® software can be used to mitigate this impact. The impacts of distributed PV, however, extend beyond the system level to the substation, feeder and even sub-circuit levels, as will be discussed in this article.
Figure 1 identifies the utility processes and systems that need to be adapted to account for the growth of PV on the distribution grid. The processes and systems impacted span long-term planning, to hourly and day-ahead forecasting, as well as from an individual sub-circuit to an entire distribution area.
As a first post exploring the adaption of utility processes and systems, we’ll dive into the processes of hosting capacity and distribution planning, and discuss how distribution engineers can use existing tools to improve their current processes.
Distribution planning and hosting capacity: It’s time to model reality
Before distributed generation, utilities modeled their grid by focusing only on the consumer load that needed to be met. Today, however, many utilities across the U.S. are focused on calculating the PV hosting capacity of their feeders, and on understanding how distributed PV may alter their existing distribution planning. Conveniently, both challenges fall in the hands of the distribution engineer and require similar knowledge of distributed PV.
Today, the distribution engineer’s tools of choice (CYME, EDD, Milsoft, Synergi and others) often do not contain knowledge of where behind-the-meter distributed PV is located, the rate at which it’s growing, or the irradiance data needed to simulate PV production.
As a result, distribution engineers are not readily able to:
- Simulate PV production at the feeder-level
- Determine differences in PV installation rates between feeders
- Simulate PV on an 8760 basis in a manner that is time-correlated with load
Without these capabilities distribution engineers cannot:
- Capture the contribution from existing PV to their load today
- Account for differences in PV installation due to customer demographics
- Understand the differences between sunny, overcast and partially cloudy days
This lack of information and capability leads distribution engineers to often apply overly conservative (but not necessarily accurate) tests when determining hosting capacity. As a result, they find themselves applying generalized PV adoption metrics when undertaking distribution planning processes. This hurts the utility and its ratepayers in two ways.
Two impacts of not accounting for distributed PV
The first way not accounting for distributed PV hurts utilities and their ratepayers is through the regulatory process. The evident short-comings of current modeling processes (e.g., failing to time-correlate PV production with system load when simulating power-flow) gives regulators reason to doubt the findings of the utility. This is dangerous for all stakeholders since the utility is responsible for the safe and reliable operation of the grid.
The second and less obvious way is through the utilities’ annual distribution grid investments. These investments are becoming progressively riskier as DER adoption and technological change continues to transform the way the grid operates. If these investments become stranded assets, cost recovery will protect the utility in the near-term, but will also lead to rate increases, exposing the utility to further DER adoption. As such, utilities and their ratepayers are dependent upon the utility making the best investments possible to minimize costs and maximize everyone’s return on investment.
Improve your processes by using your data
The good news for utilities is that the information they need to improve their processes is often being captured through the customer interconnection process. Most utilities are now aware that they should be capturing PV system size, location, orientation, tilt and components as part of their interconnection process. For utilities using PowerClerk®, this information has been captured since they started using PowerClerk.
Over the past two years, the SolarAnywhere team at Clean Power Research has been working with leading utilities to incorporate their existing distributed PV data into their hosting capacity and distribution planning processes. Using SolarAnywhere® FleetView®, our utility partners have changed the way they undertake distribution planning and how they calculate hosting capacity to fully incorporate the effects of distributed PV on a feeder-specific basis.
Want to take the first step?
We’re excited to be helping our utility partners engage with the energy transformation. Having worked with utilities across the country, we’re familiar with the challenges they face when incorporating the effects of distributed PV into their processes. The good news is that we take the knowledge gained from tackling each of the challenges, and build it back into our products.
If standing on the shoulders of your peers and incorporating the effects of distributed PV sound good to you, then we invite you to connect with us at firstname.lastname@example.org.
Learn more about solar forecasting
Access research that demonstrates how solar forecast model accuracy changes depending on the geographic footprint of the area being forecast, and evaluates the economic impact of remedying forecast errors.
The SolarAnywhere® team transforms satellite images into industry-leading solar irradiance data. We make algorithms, not satellites, so when NOAA/NASA upgrades the hardware, we take notice of the possibilities—and get to work.
On January 2, 2018 we made a seamless transition to the new NOAA/NASA GOES-East (Geostationary Operational Environmental Satellite) for real-time and forecast solar irradiance in Eastern North America. Since the GOES-16 satellite has new capabilities and each piece of hardware is unique, we spent much of November and December 2017 calibrating SolarAnywhere values from test images against data from high-quality ground monitoring stations and the GOES-13 satellite it replaces.
The new GOES-East has 16 spectral bands instead of five, and up to four times the resolution (down to 500 meters). Visible light bands produce images like the one in Figure 1 below, which shows a clear January day in the Northeast with recent snow. IR bands already enable our models to distinguish between white clouds and white snow, two very different things for the purposes of solar production. In the future, additional spectral information may be used to better capture aerosols or smoke, for example, from the fires that hung over much of the Western U.S. in 2017.
In addition, the new hardware scans the Continental U.S. 12 times per hour, a threefold increase from the previous generation. The increased temporal resolution can be used to improve cloud motion models and create higher-fidelity and more accurate solar forecasts than ever before.
Utilities and independent system operators (ISOs) in high-penetration PV territories already rely on solar forecasting to help schedule generation. Looking forward, we’re exploring the possibilities for ever higher resolution forecasts, for example as an input to the intelligent energy management system of a single commercial building. We continue to innovate through tech advancements in everything from machine learning, GPUs for processing and new satellites, as well as partnerships with forward-thinking companies and research organizations.
For the moment, our historical data is unchanged by the new satellite—obviously it will be years before images from the new satellite make up a significant portion of the archive. However, paid license subscribers to the SolarAnywhere Data or SystemCheck® APIs now have access to V3.2 time-series data up to real-time, current hour, reflecting our commitment to provide our users with the most accurate and up-to-date information possible.
Register for a free account to give SolarAnywhere a try!