For months now we have heard President-elect Joe Biden tout the job and economic growth that would come from transitioning to a clean electric grid. We’ve also heard from critics who say the new clean electric grid he is proposing will cost upwards of $2 trillion.
These assumptions about costs are misguided, as are other widely-held assumptions about a clean electric grid. Transitioning to a clean electric grid could actually cost less money and save us billions of dollars, create jobs, and result in a cleaner, more reliable grid across the United States.
We found that when you use better planning models and scale both local solar and storage, as well as utility-scale solar and wind, you maximize cost savings and unlock the path to the lowest cost grid. In fact, it could generate nearly half a trillion dollars in savings to ratepayers over the next 30 years.
In recent years, myriad studies conducted in more than 20 states have tried to evaluate the net value of distributed energy resources (DERs) like local solar and storage. While these studies have been used to influence the implementation of tariffs for DERs on a limited scale, none have been able to project those costs and benefits at scale into total system planning processes in the all-important capacity expansion and production cost modeling that underpins utility system resource planning. In fact, most grid and system planning processes aren’t equipped to consider resources based on their total costs and benefits to the entire system. That’s because they analyze the grid in a piece-meal fashion in distribution, transmission and generation modules, lack exhaustive data inputs, and can’t fairly consider smaller resources like DERs.
We wanted to know what the grid would look like, and cost, if we stopped ignoring the benefits of DERs and optimized the integration of these resources through a better modeling process aimed at a true least-cost development plan for the entire grid. So we engaged Dr. Christopher Clack of Vibrant Clean Energy to apply his advanced and big-data friendly WIS:dom(R) model to the task. What we found surprised even us.