In Q3 2024, I approved what I thought was a straightforward purchase: two ChargePoint Level 2 chargers for our parking lot. We had maybe twelve EV-driving employees at that point and the request had been sitting with operations for a year. I budgeted roughly $8,000 for the chargers and installation, submitted the requisition, and expected the work to wrap inside a month.
It took us about fourteen weeks. The final approved spend was closer to $23,000, and two-thirds of that had nothing to do with the chargers themselves. The embarrassing part: the line item I originally handed to finance had the word "charger" on it four times and "transformer" zero times.
The problem I thought I was solving
Here's what I believed at the start: EV charging is a consumer product category. You buy a box, an electrician installs it, done. ChargePoint is a recognized brand, Level 2 is the common standard, and our facility manager confirmed we "had enough power." I took that at face value. I'd read a few comparison articles that put installation between $1,500 and $4,000 per unit, which fit neatly into the budget I had in mind.
Everything I'd read framed the decision as a charger question—brand, amperage, network subscription. In practice, the charger was the cheapest and least complicated part of the project. It took me about six weeks and four vendor meetings to figure that out.
The short version of what we hit: our existing feeder transformer didn't have the spare capacity the facility manager assumed it did. Once two licensed electricians independently confirmed that, the entire project scope changed. We weren't buying chargers. We were looking at a pad-mounted transformer upgrade, new conduit runs, load calculations, a utility coordination process, and — because our CFO now wanted visibility into the whole thing — an energy-monitoring layer we hadn't discussed at all.
What the problem actually was
Some of this was my own gap. I'm an administrator, not an electrical engineer, and I approved a scope of work I didn't fully understand. But a bigger part was structural, and I think it's the reason facility energy purchases go sideways at companies our size.
The first layer was capacity nobody had measured. "Enough power" is not a number. It's an assumption. Our building was built in 2009, and the original transformer sizing had been done for the loads of that era. Nobody had documented headroom since. When we started asking vendors for quotes on a replacement pad-mounted transformer, the specifications varied wildly depending on who we asked. Some quotes assumed we'd stay at existing load. Others assumed we'd add the chargers plus 20% margin. One included future-proofing for additional chargers we hadn't asked for.
I can only speak to our situation, but I suspect this is common: the moment you ask three vendors to size a transformer, you get three different answers because the input assumptions are all different. ABB's transformer product line, for example, is broad enough that two resellers can quote "an ABB transformer" and mean entirely different units. That was a genuinely useful thing to learn — and a frustrating one, because it meant I couldn't compare numbers until I forced everyone onto the same load calculation.
The second layer was integration. Once the CFO saw the transformer quote, she asked a reasonable question: "How do we know if this was the right size?" That's how the Schneider Wiser Energy Monitor conversation started. We looked at a few monitoring options, including the Wiser line, because we already had Schneider equipment elsewhere in the building. But monitoring isn't a plug-in accessory — it has to be designed in. The CTs, the panel space, the data pathway, the question of who reads the dashboard and what they're expected to do with it. It became its own sub-project.
I should mention: at one point, while we were waiting on a utility coordination date, our facility team asked whether a portable power station could bridge a planned outage for the server room. We spent two days figuring out the SBAOH portable power station charging requirements and realized the answer was "technically yes, practically no" for a load that size. Useful exercise. Not a solution. But it did force us to actually map which circuits were critical, which nobody had done.
The third layer was the invoice and accounting trail. This is the part I care about most, and the part no vendor article warns you about. The transformer got quoted through a reseller. The installation was a separate electrical contractor. The monitoring was a third party. The ChargePoint network subscription came as a recurring software charge. Four vendors, four invoicing formats, and one of them — the reseller — sent us a handwritten delivery receipt with the wrong purchase order number. Finance rejected the $6,200 expense report on the first pass. I ate a week of back-and-forth over a typo.
I'd managed roughly $280,000 a year in spend across eight vendors before this project, and I thought I'd seen every invoicing failure mode. What I hadn't seen was a project where the invoicing complexity scaled with the technical complexity. Each vendor assumed the other vendors were handling documentation. Nobody was.
What the delays actually cost
I want to be concrete here because "scope creep" is a phrase that hides numbers.
- Timeline: 14 weeks from requisition to first charge, versus the 4 weeks I'd assumed. Two of those weeks were internal — chasing a signature and waiting for finance to reprocess the rejected invoice.
- Direct overrun: roughly $15,000 above original estimate. The transformer and utility coordination were the bulk of it.
- Advisory cost: about $2,400 for a third-party electrical engineer to review the transformer sizing, because CPAs and facility managers disagreed on the load calculation and I needed a neutral party.
- Opportunity cost: the server-room circuit mapping we did during the outage planning turned out to be genuinely valuable, but it pulled our facility lead off other projects for six working days.
To be fair, the project delivered what it was supposed to deliver. Employees are charging their cars. The CFO has dashboards she actually uses. The transformer has documented headroom for future expansion. But I approved the original requisition on the basis of a number that was wrong by a factor of three, and I'd rather not repeat that.
The short version of what I'd do differently
After five years of managing procurement, I've come to believe that the energy equipment category is where procurement discipline fails most often, because the technical language makes non-engineers defer to whoever sounds most confident. The traditional advice — "get three quotes" — doesn't help when the three quotes aren't measuring the same thing.
What worked, when we finally applied it, was a five-question pre-check before any facility energy purchase moved forward:
- What's the actual measured load, and where's the documented headroom? Not "we have enough power." A number, in writing, from someone licensed to state it.
- What's the smallest scope that gets us to the outcome? In our case, a smaller transformer with a monitoring layer might have deferred the full upgrade by two years. We never asked.
- Who signs the load calculation, and does everyone agree on it? Force every vendor onto one shared specification before comparing prices. This alone would have saved us three weeks.
- What's the invoicing path for every line item? Name the entity receiving each invoice, and confirm they can issue a compliant one before you place the order. This is the question that would have saved me the $6,200 rejection.
- What recurring software or subscription cost rides along? Charger networks, monitoring dashboards, and firmware plans are recurring. They belong in the total cost, not in a footnote.
That's it. No framework, no maturity model. Just the questions I wish someone had handed me before I typed "EV charger" into a requisition and hit submit.
This worked for us, but our situation was a mid-size company with a single building and a relatively patient CFO. If you're dealing with multi-site facilities or a denser load profile, the calculus — and probably the number of vendors — is going to be different. I can only speak to what I've actually run.
Prices referenced in this article are from Q3–Q4 2024 quotes received by our organization; verify current pricing and specifications directly with vendors. Regulatory requirements for electrical installations vary by jurisdiction — confirm with your local authority before proceeding.