A Midwest cold chain facility called us three weeks ago about a failed lithium conversion. They'd bought 31 packs from another supplier, eighteen months prior. Spec sheet looked fine. 48V, 314Ah, LiFePO4, CE mark. The procurement team thought they'd covered everything.
Fourteen packs were already below 80% capacity. The supplier had gone with 16S passive balancing to win on price. Facility ran three shifts with opportunity charging. Passive BMS can't maintain cell balance under that kind of cycling. Each charge made the imbalance worse until the weakest cells started dropping out.
New packs, expedited freight, install labor, rental trucks for three weeks while they waited. Total damage came to forty-seven thousand dollars more than if they'd specified the right configuration from the start. That's not a quality problem. That's a specification problem, and it happens constantly because procurement teams don't know what questions to ask about internal battery architecture.
The Gap Between Spec Sheets and Reality
Voltage and capacity are what most RFQs focus on. Sometimes cycle life. Maybe certifications. Internal cell architecture almost never comes up.
But a 48V battery can be built multiple ways. 15S with standard LFP cells. 16S for 51.2V nominal. 14S if someone's using higher-voltage NMC chemistry. Same headline numbers on paper, completely different behavior in the field. Different BMS requirements. Different charging profiles. Different failure modes when something goes wrong.
We've been tracking warranty data across Polinovel deployments for years now. Configuration mismatches drive somewhere around a third of claims in the first two years. Not bad cells, not BMS failures. Procurement got the wrong architecture for the application.
What Actually Happens in Series Configurations
Every cell in a series string carries identical current. Sounds like basic electrical theory because it is. The practical consequence is that one weak cell tanks the whole pack. In a 16S setup, when that one degraded cell hits low-voltage cutoff, BMS disconnects everything. Truck stops mid-aisle. Happens constantly in facilities running hard three-shift schedules.
Higher voltage does have real advantages though. An 80V forklift system pulls maybe 40% less current than 48V for equivalent power to the motor. Less current means you can use smaller cables. Less heat in the connectors. Less wear on the wiring harness over time. There's a reason heavy-duty counterbalance trucks use 80V platforms once you get above 6,000 lb capacity ratings.
Parallel Configurations Look Simple Until They Aren't
Parallel wiring adds capacity while keeping voltage constant. You'd think that makes it the easier option.
Current sharing between parallel strings creates problems that don't show up until months into operation. Even cells from the same production run have slight differences in internal resistance. The string with lower resistance naturally draws more current. Wears faster. Fails earlier than it would have on its own.
We saw one parallel pack last year where one string had logged 40% more cycles than the other. Same cells, same BMS, same truck. Different cable lengths to each string. The shorter cables had lower resistance, so that string worked harder on every single cycle. Customer found out when the worn string failed and warranty got denied because it wasn't a manufacturing defect. Installer took a shortcut.
Good parallel installations use matched cells from the same lot and equal-length cabling. Takes more time. Costs more. A lot of aftermarket battery shops skip it.
Actual Numbers from Fleet Conversions
We pulled data from 43 customer fleets that switched from lead-acid to lithium over the past couple years. The results split into two clear groups.
Fleets that matched their configuration to how they actually operate saw payback periods in the 26 to 31 month range. Battery life projections holding at eight years or better.
Fleets that bought primarily on price and spec sheet numbers projected 19-month payback initially. Real-world payback stretched past 40 months because capacity faded faster than expected.
The difference between these groups came down to balancing architecture. Active cell balancing runs maybe eight hundred to twelve hundred dollars more per pack at purchase. It redistributes energy between cells continuously, during charge and discharge. Passive balancing just bleeds off the high cells during charging. Works fine if you're doing full overnight charges. Falls apart when you're opportunity charging between shifts.
Twenty-truck fleet running two shifts. Active BMS spec adds twenty thousand to initial cost. Those packs hit 4,000 cycles before capacity drops to 80%. Passive systems under the same duty cycle get roughly 2,400. Over a seven-year window, the active fleet does one battery replacement. Passive fleet does two. Twenty thousand extra upfront saves around seventy-five thousand over the equipment lifetime.
The BMS Compatibility Issue Nobody Warns About
Modern forklifts want to talk to their batteries. State of charge on the dash. Fault codes when something's wrong. Sometimes interlock functions that won't let the truck operate if communication drops.
BMS has to speak the same protocol as the truck controller. Crown uses one thing. Toyota uses another. Yale is different again. Buy a pack that integrates perfectly with Crown trucks and it might throw constant errors on a Toyota.
We did an informal survey across customers using multiple OEM brands. Something like 30% of aftermarket lithium installations have at least partial communication mismatch. Trucks still run, so nobody complains initially. But fleet managers can't see real battery health data. Can't catch problems developing. Can't use the diagnostic tools built into the trucks they already own.
Thermal management is the other place where cheap packs fall short. A spec sheet might say the battery operates from negative 20C to positive 50C. That doesn't mean it performs well across that whole range.
Lithium cells risk permanent damage if you charge them when they're cold. Quality BMS systems lock out charging below zero until the pack warms up. Which means in a freezer operation, trucks coming out of a negative 20 zone might not accept a charge until they've sat in the warm area for a while. Operations people hate that.
Polinovel cold storage packs include heating elements that keep cells warm enough to charge. More expensive and more complex than standard packs. Eliminates the "truck won't take a charge" problem that drives freezer operations crazy when they're using batteries designed for ambient warehouses.
Matching Configuration to How You Actually Operate
Single shift operations with full overnight charging available:
Passive balancing is fine. Lower-cost 16S configurations work acceptably because the duty cycle gives the BMS time to equalize cells completely during overnight charge. Watch out if you plan to add shifts later though. Batteries specced for single-shift duty don't handle expansion well.
Multi-shift with opportunity charging:
You need active balancing. No way around it. Partial charges never let passive systems fully equalize. Imbalance accumulates cycle after cycle. Eventually you start getting protective shutdowns during operation.
Charger selection matters here too. Opportunity charging works best with high current for short bursts. Thirty minutes at break time rather than hours at shift change. If your chargers can't push enough current in available windows, trucks don't get enough energy to make it through the next shift.
Lead-acid loses a third of its capacity or more at freezing temperatures. Lithium holds up better but has charging constraints. Specify thermal management rated for your actual operating temperature, not just the coldest number you might hit occasionally. A pack rated to negative 30 doesn't need active heating in a negative 10 freezer. It absolutely needs heating in negative 25 blast freezer applications. Procurement teams using generic cold storage specs instead of facility-specific requirements miss this constantly.
Continuous operation with maybe 10 or 15 minutes available for charging between missions. The battery needs to accept charge fast, which not all lithium chemistries do equally well. If you're adding parallel capacity for longer runtime, BMS needs to monitor each string independently. Otherwise you can't catch current sharing problems before they cause failures that shut down automated operations with no warning.
What Your RFQ Should Include But Probably Doesn't
Before you send anything to suppliers, document how your operation actually runs. Specific hours per shift. Specific number of shifts. Actual charging windows between shifts. What you expect the operation to look like in three to five years. Vague language like "multi-shift operation" gives suppliers room to propose configurations that technically meet requirements but underperform real-world conditions.
Check your charger situation. Lead-acid chargers almost never work for lithium conversions without modification or outright replacement. We've seen quotes that left chargers out entirely. Looked competitive until the real installation scope showed up.
Figure out what communication protocol your forklifts use. This varies by manufacturer and sometimes by model year within the same manufacturer.
When you write the spec, require suppliers to tell you what their cell architecture actually is and why they chose it for your application. If all you get back is a restatement of the output specifications, that's someone selling you commodity product, not an engineered solution.
Ask directly about cell balancing. Active or passive. What duty cycle can it actually handle. Suppliers who know their products give straight answers. Evasiveness suggests either technical limitations they don't want to discuss or simply not knowing enough about their own products.
Get references from similar operations. A supplier with good results in single-shift ambient warehouses might have zero relevant experience for a three-shift freezer application. The technical challenges are different enough that track record in one segment doesn't transfer.
During evaluation, run TCO calculations with your actual numbers. Suppliers model favorable assumptions. Use your real electricity costs, your maintenance labor rates, your equipment utilization.
Look at the counterweight solution. Lithium packs weigh way less than the lead-acid they replace. A 48V lithium might be 600 lbs where lead-acid was over 2,000. That weight difference affects forklift stability. Good suppliers include ballast in their proposal. Others leave it as a hidden cost or a safety gap you discover after installation.
Read warranty terms carefully. Standard warranties often exclude extreme temperature applications, high-cycle operations, or installations done by anyone other than authorized technicians. A warranty that doesn't apply to how you actually operate is worthless regardless of what it says on paper.
How We Approach This at Polinovel
Engineering evaluates applications before we quote anything. Takes longer than just pulling standard product. Prevents the specification mismatches that generate warranty claims and customer complaints.
We start with operational data. Shift patterns, equipment models, charging infrastructure, environmental conditions. Model expected performance across different configurations. Recommend based on what the customer says matters most to them, whether that's lowest initial cost, longest battery life, or fastest payback. Those priorities lead to different recommendations and we'd rather have that conversation during specification than discover the mismatch two years into deployment.
Complex applications like cold storage, AGV integration, or mixed fleet conversions get site assessments. Technical staff comes out, looks at infrastructure, watches operations, develops specifications that account for how things actually work rather than how they're described in emails. Those assessments have value even if the customer ends up buying from someone else because they generate documentation useful for evaluating any supplier.
Standard products cover most material handling applications. FL51 series for 48V systems, capacities from 314Ah up to 920Ah depending on equipment requirements. FL80 series for 80V heavy-duty applications. Golf cart, AGV, specialty vehicle configurations for those segments.
We treat specification as an engineering problem. Sometimes that means telling customers our products aren't the right fit for what they're trying to do. Better to have that conversation upfront than deal with a failed installation later.
Next steps if you're evaluating options:
Fleet assessment from our technical team documents your operational parameters and models expected performance across configurations. Delivers a specification document useful for competitive bidding regardless of who you ultimately buy from. Typical turnaround runs five to seven business days for standard applications.
Configuration Decision Guide covers the specification details that separate successful conversions from problem installations. Includes the procurement checklist above plus TCO modeling templates.
Technical questions about specific applications or equipment models go directly to engineering.
