Skip to Content

A Cloud Speed Sensor for PV Generators and Utility Operators

Detailed here is a sensor array intended for large (10s or 100s of MW) PV or solar thermal power plants to aid in the balance of supply and demand on an utility electric grid.
Balancing supply and demand on the grid occurs as follows:
  1. Unit commitment is conducted a day ahead based on 12-36 hour load and solar forecast. This allows slow plants to startup in time.
  2. Load following or intra-hour dispatch occurs to make up for deviations in the day-ahead unit commitment. In this way operators correct for the majority of day-ahead forecast errors.
  3. For shorter time scale deviations, automated governor control (AGC) is used to speed up or slow down turbines in thermal generators responding to frequency deviations on the grid.


Reserves are the more critical application of the cloud speed sensor. Reserves are backup power that is ordered and paid to be in standby mode just in case large forecast errors or ramps occur what would trigger (2) and (3) above. Reserves are both measured by how much energy they can provide (capacity), but also how fast they can ramp up or down to that energy level (ramp capability). The cloud speed sensor given here would provide information on the necessary ramp capability. The reserve requirements are updated on about an hourly basis so a system operator could take action within that time.
Currently this is a human based decision, but operators are already overwhelmed with information, so the true value of the cloud speed sensor will likely be realized when these decisions become more automated.
It is important to distinguish that the objective is not to detect individual clouds or respond to them, but rather to get a general idea about given current meteorological conditions (partly cloudy, cloud thickness) and cloud speed, what maximum ramp rates of the solar power plants can be expected.
This technology could be implemented as an "add-on" to the already required meteorological monitoring mandated by system operators.

Patent Status

Patent Pending
Tech ID: 22663 / UC Case 2012-120-0