How Does an Independent System Operator (ISO) Run the Electric Grid?

by Bob Shively, Enerdynamics President and Lead Facilitator

Over two-thirds of North American power is scheduled through an Independent System Operator (ISO), yet in doing seminars for Enerdynamics we often find that participants don’t really understand how that works.  

The scheduling process occurs daily as the ISO determines supply resource hourly schedules for the next day using least-cost optimization given supply offers, system conditions, and load bids. Then in real time, the ISO ramps supply up or down to balance system loads using flexible resources. Here's how it works:

A typical day at an ISO

To prepare for the next day, the ISO must schedule supply across the coming 24 hours. This is handled by the day-ahead scheduling process. Supply is scheduled hourly, meaning that providers receive 24 schedules — one for each hour of the following day. The schedule tells each provider which services it is expected to provide and the amount of MWs associated with each scheduled service. For instance, for a given hour a power plant with a capacity of 200 MW might have 150 MW scheduled to provide energy and 50 MW scheduled for spinning reserves. So for the given hour, the power plant is expected to provide 150 MW of energy and must also be prepared to ramp up an additional 50 MW if called on by the ISO.

The day-ahead process works as follows:

• At a specified time in the morning each unit scheduling coordinator submits an offer to the ISO. This offer includes services the provider is willing to provide at a given price as well as various operational characteristics such as start time, ramp capability, willingness to be started and then stopped, and minimum and maximum run times.
• Simultaneously, Load Serving Entities (LSEs) submit hourly bids for buying energy to supply loads. Although LSEs sometimes state an unwillingness to buy energy above a certain price level, in most cases LSEs simply state a forecast load plus a willingness to pay whatever the price is to receive supply.
• The ISO inputs expected system conditions including transmission and variable resource availability and then runs optimization software to determine the least-cost dispatch available to serve the required loads given the various supply offers.
• By the early afternoon, the ISO sends out schedules to each provider that was selected in the optimization process and notifies others that they have not been scheduled. The ISO also notifies all LSEs of the amount of load that has been scheduled to be served. The optimization model also creates day-ahead energy prices for each location on the grid and zonal prices for ancillary services. The energy prices are called Locational Marginal Prices, or LMPs, and reflect the actual marginal cost of serving load at each location.
• Since day-ahead schedules are considered firm, supply providers are paid the LMP for their scheduled output and loads are charged the LMP for their scheduled deliveries regardless of what actually happens in real time.

A separate group at the ISO runs the system to ensure that supply and demand are kept in balance in real time. This process works as follows:

• The real-time operators have various sources of supply available to ramp up or down given the resources selected in the day-ahead scheduling process.
• As the supply-demand balance fluctuates, the ISO ramps supply as needed based on the offers accepted in the day ahead. A real-time LMP is determined based on the price for marginal supply used in ramping.
• After the hour, the ISO calculates each unit’s actual output compared to the output scheduled in the day-ahead. The real-time LMP is applied to any deviation from what was scheduled and each provider is paid or charged (depending on whether it supplied more than scheduled or less than scheduled) based on the LMP at its location.
• Similarly the ISO compares each LSE’s schedule at specific grid locations to the LSE’s actual usage. Again, any deviation from the day-ahead schedule is either charged or paid depending on whether the LSE used more or less than scheduled.

The result is that all resources that provided supply and all loads that purchased energy receive a settlement based on their day-ahead schedules and their performance in real time.

As more and more smaller resources including demand response, distributed generation, and battery storage come onto the grid, it will be important for all involved to understand how the ISO scheduling process works. In fact, many in the industry believe we may eventually have similar distribution-level markets run by an organization called the Distribution System Operator or DSO. 

Want to learn more about how electric markets work? Enerdynamics offers its popular Wholesale Power Markets course in both live and online formats. Visit www.enerdynamics.com for more information. 

Back to Energy Insider