In the race to modernize energy storage, lithium iron phosphate (LFP) systems like HomeGrid’s Stack’d have brought undeniable advantages over lead-acid: higher efficiency, deeper discharge, longer lifespan, better energy density, and none of the maintenance headaches. But along with those benefits comes a new set of vulnerabilities that rarely get the spotlight.
One of the biggest is the “redundancy” story: the idea that a stack of independent battery modules gives you built-in reliability. The reality is more complicated.
The Redundancy That Isn’t
Each HomeGrid Stack’d module has its own electronics, communications interface, and protection circuitry. On paper, that sounds like a recipe for resilience. In practice, it’s the opposite: if any single battery in a stack stops working, it takes down the entire stack.
When that happens, you’re not just swapping a box and moving on. You have to:
- Remove the bad battery from the stack.
- Reprogram the DIP switches on the remaining batteries.
- Reconfigure the stack’s BMS so it knows its new size.
Until you do all that, the whole stack is out of commission. That’s not resilience — that’s a single point of failure dressed in modular clothing.
The Parallel Stack Problem
This fragility gets magnified when you run multiple stacks in parallel. One bad battery in one stack doesn’t just affect that stack — it throws an error across all stacks.
At that point, you may need to:
- Reprogram the master BMS to bypass the faulty stack, or
- Remove one battery from every stack to keep them matched for proper balancing.
Neither option is quick or painless. And if you’re off-grid, every hour of downtime matters.
Diagnostics: Downtime Required
Here’s where it really stings. To diagnose the problem, you have to connect a computer directly to the BMS of the affected stack with the faulty battery still in the communication chain. That means your system is offline during the process — and diagnosis can take hours.
For an off-grid home relying solely on that power, that’s a serious operational risk.
Old School vs. New School Reliability
This isn’t a HomeGrid-only issue. It’s a side effect of the entire new generation of “smart” LFP battery systems. By integrating sophisticated electronics, firmware, and communications, we’ve gained efficiency, safety, and intelligence, but we’ve also introduced more failure modes and more interdependencies.
This isn’t limited to HomeGrid, of course. Other battery manufacturers have similar issues. For example, Pytes 48100 batteries each have their own BMS. There is no “master” BMS that goes down when one of the batteries fails. However, if you have the batteries in a communication loop, which you should in most cases, if one battery fails, the entire communication loop of up to 16 batteries will stop working (and you will get to hear the annoying alarm tone from the primary battery in the loop). At least you don’t have to fiddle with dip switches to bypass the bad battery but you do need to reconfigure the communication wiring.
Lead-acid systems, for all their inefficiency and maintenance, had brute-force simplicity. A bad battery could be isolated with a breaker or fuse, and the rest of the bank kept running. No DIP switches, no firmware reprogramming, no system-wide lockout.
Why Not Open-Loop
The communication protocols we are talking about are referred to as closed-loop communication. With HomeGrid, you don’t have the choice – a stack of batteries is inherently a closed-loop system. Each battery must have an “address” in the communication loop. Technically, you could run parallel stacks without them communicating with each other, but if you want your batteries to “talk” to an inverter, this is not a good option.
Other “stackable” batteries that each have their own BMS, like Pytes, EG4, SOK, Simpliphi, and many others. These can be operated in open-loop, meaning they don’t communicate with each other. The downside of this is that the batteries can’t do balancing across the individual units, and they can’t communicate with an inverter, which means the inverter needs to work in “voltage mode.” This is not ideal with LFP batteries, because voltage is not a good indicator of state of charge, and voltage drops off dramatically when they near full discharge, which can cause a whole host of issues.
However, using open-loop does provide a level of redundancy and resilience.
Closed-Loop with Resilience
Some batteries use a different communication logic that doesn’t shut down the whole stack. For example, I have a client with EG4 LifePower4 batteries that are all communicating in two chains of 12, and combined with a Communication Hub. All of the batteries talk to each other and report the state of charge to the Hub. The Hub aggregates the state of charge. If one battery has a fault or is turned off, the rest keep on working, even though that “address” is not found in the communication loop. That is the way it should be! +1 for EG4.
The Midnite PowerFlo batteries work similarly and have the added benefit of no dip switches. They just figure out who their neighbors are and talk to each other seamlessly. If one of the batteries decides to go on vacation, the block party still goes on!
The Balanced View
On balance, LFP is still a huge step forward — cleaner, safer, lighter, and more versatile than lead-acid. But “redundancy” without true resilience is a weak form of reliability. If you’re designing a critical power system, especially for off-grid, you need to think beyond marketing claims and plan for what happens when a module fails.
In other words:
- Scalability ≠ resilience.
- Modularity ≠ redundancy.
- Smarts ≠ simplicity.
- Without the ability to keep running during a fault, LFP batteries providing redundancy is just wishful thinking.
