This installment is way overdue, but better late than never!
In the third and fourth parts of this series, I provided detail about the solar panel array that we planned to have installed on the roof of our new home. In this post I’ll describe how we structured the decision process for the Battery Energy Storage System (BESS) portion of the installation, and which one we picked. Everything is installed and operating now, so the next installment will be an update about the installation process and how things are working so far.
Let’s recap the original goals for our solar+battery system:
- Have enough solar capacity to reach annual “net zero” on our electric bill
- Have enough battery capacity to provide a safety buffer in the dead of winter
- We want to be able to run the heat pump plus a handful of critical circuits
- Have the ability to integrate V2G/V2H from an electric vehicle in the future
It turns out that fully meeting all three of these goals is, as I suspected, not feasible right now. Specifically, it’s not cost-feasible to put in enough battery capacity to run our heat pump system indefinitely in the lowest solar-generation time of the year. The best we can do in a grid outage right now is keep everything else going, and figure out how to spot-heat to keep reasonably warm. In the winter time if it’s really sunny out we can run the heat for very short periods, but that’s about it. From late spring to early fall we should have plenty of solar generation to keep everything going, even the A/C. It helps that the climate here is mild in the summer, only rarely getting above 80ºF.
As we looked at different options, we saw that each had its own set of drawbacks. We had to decide what was most important to us and evaluate the options against our prioritized criteria. We ended up loosely ranking each solution across the following dimensions, in decreasing order of importance:
- Cost
- Performance against our power needs, especially during low solar generation periods
- Simplicity of operation
- Implementation complexity and physical footprint
- Upgradeability and openness of the platform
We short-listed four options - one option that is solar-only, and three options that include BESS for backup during grid outages. The three BESS options use components from EcoFlow, Enphase, and Tesla Energy, respectively. Below is a summary of what we concluded about each option, and how we rated each of them on our highly-subjective suitability scale.
Solar Only
As someone who had to prepare untold numbers of Major Decision Framework decks over the course of my career, I obviously had to consider the “do nothing” option. We knew that we absolutely wanted to get the solar panels, as these are a no-brainer even up here in the way north. With a 30% Federal Income Tax Credit (FITC) on the cost of the system and net billing from our utility provider, the payback period is roughly 12 years for us. I look at putting in solar panels as essentially pre-paying our electric bill for the next 25 or so years, at a rate that gets better with inflation.
The financial benefit of the BESS portion is not as straightforward. In locations that have Virtual Power Plant (VPP) or Time-of-Use (TOU) rate plans, you can offset the cost of your BESS through grid feedback and/or intelligent coordination of battery use timing. We have neither of these in our PUD, at least for now. In our case, the BESS portion is effectively an insurance policy against grid outages. By having a BESS we can create a microgrid during grid outages and continue life somewhat as normal, but when the grid is up the BESS provides us no real benefit.
Going with a solar-only option would by far have been the cheapest solution and the simplest to install and operate, but it does nothing for us if the grid is down. In fact, being grid-tied means that if the grid is down the solar has to be disengaged, stranding our solar generation capability. We decided that having the insurance policy against grid outages was important enough to us that we were willing to spend the extra money to have it.
Cost: 👍
Power Needs: 👎
Operational Simplicity: 👍
Implementation Complexity and Footprint: 👍
Upgradeability: 🫳
Overall Suitability: ★☆☆☆☆
EcoFlow Technology DELTA Pro Ultra
There are several companies that make portable solar and battery energy storage systems, and EcoFlow is consistently in the list of best-rated and most popular. In the last few years EcoFlow has begun to branch out further into residential-scale systems. Most recently they released their DELTA Pro Ultra series, which has serious capacity and output capabilities. In conjunction with the Pro Ultra, they also have a residential sub-panel/automatic transfer switch (ATS) solution in their Smart Home Panel 2 product. Together, you can create a backup system for your home as small as 7kW/6kWh of capacity (kW of power output, and kWh of storage capacity), as large as 21.6kW/90kWh, or somewhere in between. The inverter and battery units are on wheels and portable, so you can take them with you if you move.
The Smart Home Panel provides inputs for the Pro Ultra units, supplying 30A/60A/90A of 240V backup power (using one, two, or three Pro Ultra units). The Smart Home Panel gets wired into your electrical system as a sub-panel, and any electrical loads that you want to have backed up during a grid outage need to be re-wired into it. When the grid is up, the Smart Home Panel feeds power to the Pro Ultra units to keep them charged. When the grid goes down, the Smart Home Panel automatically switches to battery power and keeps your critical circuits running.
We really liked the fact that the Pro Ultra units are modular, portable, and eligible for the 30% FITC. We were thinking that if we went this route, we could start with one inverter unit and 2-3 battery modules, and add more inverters/batteries as we determined how much we’d really need to cover our usage. As an added bonus, we could use one inverter and battery module(s) as a camping companion when we travel.
As we explored this option further, we uncovered serious downsides that ended up ruling it out. First, the Smart Home Panel is only rated for 100A of capacity, with just 12 breaker slots. Combining this constraint with the fact that it has to be a sub-panel meant that we would be very limited in what we could back up during a grid outage. Second, and more critically, there is no easy way to connect grid-tied solar into the EcoFlow system to charge the batteries when the grid is down. If you don’t mind some complexity, the grid-tied solar could be set up with limited microgrid capability to provide one usable outlet to charge the batteries, but that would mean extra cost and having to connect the Pro Ultra units with extension cords. That seemed more of a hack than we cared to create.
This approach could be incredibly useful in the right circumstances, but for us it just didn’t tick the right boxes to make the cut.
Cost: 🫳
Power Needs: 👍
Operational Simplicity: 👎
Implementation Complexity and Footprint: 👎
Upgradeability: 🫳
Overall Suitability: ★★☆☆☆
Enphase Energy IQ 5P
Enphase is one of the big names in residential solar, and for good reason. They have solid engineering, great features, a good reputation among installers and owners, and they’re based in the US. Not long before we began to explore our options Enphase released a major update to their BESS product line, the IQ 5P battery system. This update was a big step forward in capability, with improved storage density and power output. If you search YouTube you’ll find many videos extolling the virtues of the Enphase ecosystem.
The primary differentiator of the Enphase product line is their use of microinverters rather than monolithic or centralized inverters, making their solar and BESS systems AC-coupled rather than DC-coupled. In an Enphase system, each solar panel has its own microinverter that converts the panel’s DC output to AC. The AC output of the panel microinverters is then fed to a Combiner, which manages aggregating all AC inputs together for use in a grid-tied/microgrid system. Similarly, Enphase batteries have embedded microinverters that convert the energy stored in the batteries from DC to AC, which is fed to the Combiner as well. This is less efficient than a DC-coupled system since there are conversion losses, but the efficiency hit is not so much that it would be a deal-breaker. The big benefit is increased resiliency of the system. The inverter function is distributed, so there is no Single Point of Failure (SPOF) as in a centralized inverter system. If a microinverter fails, you lose the use of one panel or a portion of one battery module, whereas in a central inverter setup you can lose whole panel arrays, batteries, or even the entire system if the inverter fails. As an added benefit each microinverter is independently monitorable, so you can see fine-grained details of panel and battery input/output.
The Enphase IQ 5P batteries are rated at 5kWh of storage capacity and 3.84kW of power output. To get more storage and power output, you add more 5P battery modules (up to eight 5P units total). We determined that to meet our minimum requirements we’d need at least two 5P units. Our preferred setup would use four 5P units, providing us with 20kWh of storage and 15.36kW of power output. A system of that size is big enough to provide us with whole-house backup, which makes the installation easier since a sub-panel isn’t needed. The only caveat was that our heat pump system would have to be wired through load-shedding relays so that by default the HVAC would be disabled in the event of a grid outage.
The two biggest drawbacks to the Enphase system are cost and physical footprint. Our preferred design came out almost 20% more than the next most expensive option. On top of that, the Enphase system would have seven components to accomodate on the garage walls - four 5P battery units, the Combiner, the Gateway, and the load-shed relay enclosure. That’s a lot of real estate, and it would crimp our ability to use a sizable chunk of wall space in our garage.
The Enphase system is a premium product to be sure. It has features that make is very reliable, easy to use, and expandable as needed in the future. It has generator integration capability, you can connect other AC-coupled battery modules if you wish, there is an integrated EV charger available, and V2H gear is coming next year. You do pay for the privilege though.
Cost: 👎
Power Needs: 👍
Operational Simplicity: 👍
Implementation Complexity and Footprint: 👎
Upgradeability: 👍
Overall Suitability: ★★★☆☆
Tesla Energy Powerwall 3
In contrast to Enphase, Tesla takes a different approach to solving the challenges of BESS integration with your solar array. Tesla’s Powerwall-based system is fully DC-coupled, and the solar-to-grid efficiency is a fantastic 97.5%. In a Powerwall 3 system the inverter is incoporated into the Powerwall. Your solar arrays are wired directly to the Powerwall, feeding DC into the battery with no conversion losses. The Powerwall inverter provides AC power (from the solar array, the battery, or both) to the Tesla Gateway, which is wired into your grid connection. If you have more than one Powerwall, your solar array is split into multiple segments and the outputs distributed across them for balance. The upside of this setup is simplicity. The solar arrays go the Powerwall(s), the Powerwall(s) feed the Gateway, and the Gateway manages the microgrid.
The Powerwall 3 was released just this past spring, and it has specs that best its competitors for storage capacity and power output from a single module. Each Powerwall 3 stores 13.5kWh of energy and has an inverter that can supply 11.5kW of continuous power output. If more capacity or output is required, multiple Powerwall 3s (up to four) can be stacked together, and multiple stacks can be installed if needs demand. Tesla will soon release a battery-only add-on unit for the Powerwall 3, which will add storage capacity but not increase power output.
For our purposes, the minimum design was a single Powerwall 3 unit. We had a 2-unit design quoted as well. These designs provide either 13.5kWh/11.5kW, or 27kWh/23kW. From an equipment standpoint, we would have to accomodate a Gateway unit (recessible into the wall), one or two Powerwall 3 units, and a load-shed relay enclosure. Either design is capable of whole-house backup, with the same caveat as the Enphase of the HVAC being on load-shed relays.
The Tesla solution is definitely better value-for-money than the Enphase solution, regardless of the size of system installed. The quoted cost of a two-Powerwall system with twenty-three solar panels was almost 20% less than the four-battery Enphase systems with the same number of panels. The single-Powerwall quote was nearly 40% less.
The lower cost is great, but a Powerwall-based systems is not without downsides. For one, the inverter is a SPOF. If the inverter dies, you have no battery and no solar generation capability. You’re dead in the water until it can be repaired, which can be a very bad thing if the grid goes down in the interim. Additionally, Tesla is not well known for being open and interoperable. There is no generator integration available, the EV charger integration is Tesla-only right now, and V2H is limited to Cybertrucks. In summary, simple and lower cost, but constrained.
Cost: 👍
Power Needs: 👍
Operational Simplicity: 👍
Implementation Complexity and Footprint: 👍
Upgradeability: 🫳
Overall Suitability: ★★★★☆
Drumroll please - and the winner is…
After weighing all the pros and cons, we decided to go with the Tesla Powerwall-based solution. It represented the best value for dollar, took up the least space in the garage, and at least has capacity upgradeability if we need it in the future. For the price differential we just couldn’t rationalize the Enphase system, especially since most of the extra cost comes from the battery portion of the design (which, as noted above, has no payback). We weren’t really happy spending money on a Tesla product, but at the time of the decision the worst of what we’d come to learn about Tesla’s principal stockholder wasn’t yet known. If we had to make the decision again now it would turn out differently, but it’s water under the bridge now.
In the next post I’ll go over what the installation was like, where everything ended up, and initial observations about how things are working out.
Other posts in this series:
Comments