Instead of Matching Supply to Demand, Why Don’t We Match Demand to Supply? (Part One)

If we can manage electricity demand and make it a flexible resource, we open the floodgates to integration of renewables, expanded adoption of distributed energy resources, higher customer satisfaction, a cleaner environment, and more cost-effective energy distribution and consumption. The question our industry has been unable to answer is how to harness flexible demand and use it effectively to reap its benefits.

The quest to know this piece of the puzzle is what brought me to work at Tendril – to have the chance to develop the technology utilities need to make flexible demand management a reality. As the Economist recently reported, we need this technology to integrate renewables in economically sound ways that balance supply and demand; it’s technologies like digitalization, smart meters, and batteries that enable households to flatten their peaks, to shift demand to different parts of the day, to manage intermittent supply, and to store energy more efficiently.

As I’ve worked to develop this technology at Tendril, I’ve also learned that what works on a technical level to solve a demand problem will only be successful if it also works on a human level. Customers have to be comfortable, physically and economically, with their energy management choices. That’s where a lot of our experimentation lies now: How do we talk to consumers about flexible demand? How do we engage them in it?

The bottom line is, we can’t match demand to supply without customer participation. And the first step to participation is making home energy intelligence feasible: customers don’t want to think about energy or comfort; they want it to all work automatically behind the scenes. When it’s not working, however, they want a way to easily communicate with their utilities to fix the problem. With that knowledge of customer preference in mind, let me backtrack a bit and share some of what I’ve picked up along the way about flexible demand and the potential it holds.

Unprecedented Energy Savings Out West

In 2013, I was working at the National Renewable Energy Laboratory (NREL), where I published with my NREL and Lawrence Berkeley National Lab (Berkeley Lab) colleagues, the first technical and economic study of the type of demand that is not just responsive to a price signal, but is actually flexible. The flexible aspect of demand refers to the load activities, like heating and cooling your house, that can be re-scheduled. A simple example is scheduling tasks like running the dishwasher during the least cost hour of the day. Or, managing the temperature or comfort of the home while also optimizing the time of day that the HVAC system operates. The study provided a quantification of how much load across various sectors (residential, agricultural, commercial, etc.) we could remove from one slice of time in the day and move to another slice of time. We calculated the kilowatts of load available in any given hour across the Western Interconnection -- a wide span of our electric grid from western Canada south to Mexico, east to the Great Plains and west to the Pacific Ocean.

After determining available load capacity at any given time, we tackled the real hard question: what would happen if you offered that flexible load to a wholesale market? We took a model of the entire Western Interconnection that accounted for all generator and transmission constraints and we plugged in demand as a flexible resource, creating the first model of time varying storage capacity. We ran the system twice -- with and without flexible demand -- to quantify the impact of flexible demand on the whole system. We sought answers to these questions: Would we avoid any natural gas or coal generation? Did we serve load more reliably? Were transmission lines less constrained? And, most importantly, how much would the market pay for flexible load?

The results were consistent with our studies on battery storage technologies, but being able to quantify flexible load value allowed the industry to compare one flexible load technology against another. Overall, dispatching flexible loads saved conventional energy resources (flexible demand created 113 MW of capacity and shifted 135 GWh of energy across the Western Interconnect) and served load better (we found flexible demand can meet about 33% of the need for frequency regulation, 19% of spinning contingency reserve, and 85% of flexibility reserve). There was clear societal benefit to implementing flexible demand.

To study the marginal price that flexible load cleared in the market, we modeled a subset of the system: Colorado’s balancing areas of Public Service Colorado (PSCO) and Western Area Colorado Missouri (WACM). In this market, all of the demand response resources combined had a production cost value of about $27/kW-y. The total value, capacity plus production, is in the range of $60/kW-y to $250/kW-y. But amongst the flexible demand resources, the value varied considerably and was driven by two factors: the total flexibility (how many kW can be re-scheduled how many hours away from their original time) and the coincidence of load flexibility with system needs (how many kW are available at the peak load hours of the system). When we constrained residential cooling loads to only reschedule 1 kWh, which is about 15 minutes of cooling system operation, between 6am and 6pm, the production value was only about $5/kW-y. At Tendril, we have investigated and field tested the flexibility of residential cooling and found that on average we can reschedule about 18 kWh between 6am and 9pm That increase in flexibility is worth about $35/kW-y to $50/kW-y. Flexibility of demand also lowers the cost of integrating more wind and solar generators into the system.

With the benefit determined, the next step was to determine whether flexible demand is technologically and socially feasible. That’s what I came to Tendril to explore: to find that flexibility in residential load that is a net benefit for society, meaning the consumer benefits from a better experience and cost savings, and the utility gains flexibility at the grid edge, where it needs it most.

Continue to my next post to read about the impact of decarbonization in California and the work my Berkeley Lab and NREL colleagues have pursued integrating flexible demand into California’s energy landscape.


1M. Hummon, et al. Grid Integration of Aggregated Demand Response, Part 2: Modeling Demand Response in a Production Cost Model. National Renewable Energy Laboratory. NREL/TP-6A20-58492 December 2013.





  • Continuous Demand Management
  • Customer Ops & CSAT
  • DERs
  • DSM
  • Data Analytics
  • Demand Response
  • Disruption
  • Energy Efficiency
  • HERs
  • High Bill Alerts
  • Privacy & Security
  • Real Estate
  • Smart Home
  • Solar

    Subscribe to This Feed