What does "balancing the load" mean and why is it important for a sustainable energy system?

The renewable energy industry currently faces a big challenge - energy storage.  Solar energy is currently the fastest-growing energy source and one of the most promising methods of renewable energy, but unlike other energy sources, the amount of power generated fluctuates with the sun.

To illustrate the energy storage challenge, we can look at a chart of net electricity usage throughout a 24-hour period:

They call this the “duck curve” as the shape of the curve can (kind of) look like the body shape of a duck. The lowest point in the middle of the chart (the body of the duck) is when there is peak solar generation in the middle of the day. The highest point (the head of the duck) is where there is peak demand. This time is right after sunset - when solar energy is not generating, people come home from work, turn on lights and make dinner.

The chart shows that as more solar generation is added to the grid, the more extreme the highs and lows become. If energy can’t come from solar during these peak demand times it has to be generated by a power plant that can be spun up for these peak load moments. These “peaker” plants are typically powered by natural gas - so, if our ultimate goal is to eliminate carbon emissions, we not only need to generate power from renewable means - we need to find a way to "balance the load" without relying on peaker plants. Balancing the load will likely mean that we need to find ways of storing the power generated from solar during the day to be able to use it once the sun goes down.

The One-Week Design Sprint

Inspired by the 1-week sprint structure from the IDEO Co-Lab we first used this challenge as a way to practice this way of working. A synopsis of the week is as follows:

Day 1: Understand the Brief - A deep dive into what we will be working on to learn as much as we can about the problem.

Day 2: Inspiration, Sketching, and Build Plan - The day starts with desk research into the challenge, including other solutions and analogous solutions from other industries. In the middle of the day, we make sure to sketch ideas that were inspired by the exploration, and at the end of the day, we gather to discuss what we might build the following day.

Days 3 & 4: Prototyping and Iteration - For two days we worked to build a proof-of-concept of the idea, then iterated on the concept in preparation for the share on the final day.

Day 5: Documentation and Sharing - On the last day we documented and shared the idea with our colleagues.

When we got into sketching, there was a clear theme around managing heat or using heat to store energy.  After looking into some ideas around using a home as a thermal mass to lower peak energy usage, and all-in-one solar refrigerators, we settled on one idea to dig into further: 

A refrigerator that freezes ice when energy is plentiful - and uses the ice for cooling when energy is scarce.

We called it “The Grove” because we estimate that using a load-balancing fridge could have an equivalent carbon impact as planting and caring for a grove of 6-7 trees.

Some initial mock advertising materials and block diagrams from the one-week sprint.

Our Initial Questions

The idea seemed sound - a standard home refrigerator/freezer already makes ice, so we imagined these mechanisms could be repurposed to store energy in ice as well. However, there was more we wanted to learn and to get there we would need to do some prototyping and modeling. At IDEO we commonly refer to prototypes as “embodied questions.” The questions we wanted to answer were:

1. What ballpark volume of water will we need to keep a refrigerator cold?

2. Is the daily frequency of energy availability, storage and usage compatible with the concept?

3. What might the real carbon impact be?

4. What might be the business case for selling this new type of refrigerator? Would it pay for itself?

5. What kind of functional prototype could we build in two days, in the early stages of COVID-19 lockdown without having to leave our own homes?

Data Modeling

I have solar panels on my roof and have been using the Sense energy monitor for a few years. This allowed us to have a real dataset to dig into and look for opportunities.

Looking at my data - a day looks just like the “duck curve” problem. An energy peak at noon and peak demand right when the sun sets. Even though I’m generating solar power, my energy could be coming from fossil fuels at night.

My energy monitor classifies my biggest users of electricity. The biggest constant users of electricity are my freezer and my fridge:

We ran the fastest experiment we could. Let’s have a mini-freezer turn on when solar generation is above the power it needs to run, and then turn off when solar power is not available. Can we keep a 6-pack cold by freezing ice only with solar power?

Temperature sensor in a mini freezer - the freezer turned on only when solar was available. When the temperature got too hot in the cooler, ice was moved from the mini freezer to the cooler

Week 1 results

• The ability to transition to intermittent power draw makes the fridge harmonious with solar power availability (93% vs. 25% for a regular fridge)*

• Potential carbon savings equivalent to 6.5 trees*

• Prevents the release of 2 tonnes of CO2 over its lifetime

• Translates into up to a $850 investment into the global economy of future generations**

* Estimate based on my actual fridge use and solar generation over ten months
** Using an estimated 50lbs of CO2 per tree per year https://www.ecomatcher.com/how-to-calculate-co2-sequestration/
*** Assuming a social cost of carbon of $417/tCO₂ https://en.wikipedia.org/wiki/Social_cost_of_carbon#cite_note-JCP2019-12

Higher Fidelity Engineering and Prototyping

To help answer some of the remaining technical questions and to understand how an energy storage refrigerator might work, we began a second sprint project.

Following up on the previous sprint, our goal is to understand what kind of technology is needed to maintain a refrigerator at temperature overnight without power, until solar energy kicks in the next day. To make use of solar resources in the daytime, the system would store energy by freezing during the day and cool the refrigerator without running the compressor when no sun is shining. 

We started by looking into some fundamentals about refrigeration and heat transfer:

• How would we design/augment a refrigerator's cooling system?

• How much ice would we need? How much energy is needed?

• What are refrigerator/freezer set points? What is the temperature of the evaporator coils?

• How much heat energy is absorbed when melting ice - the heat of fusion - and how that compares to a liquid that isn’t frozen but just cold

We ran some tests to quantify the amount of energy that would need to be removed over an 8-hour time period. We found some temperature test data for refrigerators that were unplugged and used that to calculate the energy of the system before and after. This analysis requires assumptions of the heat capacity of items in the refrigerator and is so dependent on this value that our range of ice volume needed was so large, that it didn’t tell us much.

An ice-based cooling system makes sense for a refrigerator, but the freezer is more difficult. The set point temperature of the freezer is around 0F quite a bit colder than the fridge. It would be possible to cool the ice below 0 and keep the freezer cold for a while, but since the ice wouldn’t melt, only a small amount of energy could be retained by the ice.  

A refrigerator with a freezer cools down the freezer and uses the freezer air to cool the fridge, so using regular ice just won’t work - we will need to use a different substance where we can set the liquid to solid phase-change of the material below the temperature we would keep the freezer.

With this background research complete - we built a prototype heat exchanger.  We didn’t have ready access to a freezer that could freeze the glycol water mixture at very low temperatures, so we used regular water and tested it at refrigerator temperatures:

36 water-filled copper pipes with end caps on the bottom, wrapped in insulating foam. A fan was mounted on the side to draw air across the pipes

We tested our prototype in a full-sized refrigerator. The heat exchanger was frozen and placed into the unplugged fridge. With only ½ gallon of water, the system kept the refrigerator cold for over 7 hours:

This proof-of-concept gave us the confidence that the idea is worth pursuing.

A Design Iteration

From here, we wanted to talk to an expert in the field. We called up Kipp Bradford - CTO of Heat pump company Gradient and former senior Research Scientist at the MIT Media Lab. He’s been living and breathing thermodynamics for quite some time.

He thought we had something interesting but pushed us to consider how we might go further with the idea to make it more appealing to customers and potential clients.

Next week we'll publish the next article of this series and share where this exploration took us.