Telecom equipment typically draws 50 to 70 percent of its stated nameplate load. Most backup power systems get sized off that nameplate anyway, which means a real number that should be part of the math never makes it into the spreadsheet.
Evoltix builds the Zero-Glitch Power Module, a power intelligence system, and one of the more common conversations Bernie Dugan has with prospective customers has nothing to do with the ZPM at all. It is about the system they already have, and why it costs more than it should.
How a System Gets 25% Bigger Than It Needs to Be
Sizing a backup power system usually falls to whoever owns the site’s IT infrastructure, not a power engineer. Dugan describes the typical process: the team either sizes off of what was previously installed, or, if there is a more sophisticated support department, off a list of gear that is slated to go on the system. Both approaches use the equipment’s stated draw, the number on the spec sheet, as the baseline.
The problem is that the number on the spec sheet is a ceiling, not an average. Telecom gear runs at 50 to 70 percent of that stated draw under normal operating conditions. A system sized to the spec sheet instead of the real draw ends up with more power conversion capacity and more battery capacity than the site will ever actually use, adding up to a system that is 25 percent or more larger, and more expensive, than the load requires.
None of this is a mistake in the sense of someone doing their job poorly. It is what happens when the only number available is the one printed on the equipment, and nobody has a reason to question it until the invoice arrives. A sophisticated support department with a documented equipment list is, if anything, more likely to fall into this pattern than a smaller operation guessing off an old installation, because a documented list feels like due diligence even when every entry on it is still the nameplate figure.
Nobody chose to overbuild the system. Everybody just used the number that was easiest to find.
What a Load Study Actually Measures
The fix is not complicated, and it is also not something most sites have ever done. A load study measures real energy draw at the site over time, rather than assuming the spec sheet tells the whole story. Depending on the application, it can run as short as a day for a fairly static server room, or as long as a month for a power site with more variable demand. Evoltix deploys a monitoring unit at the site, retrieves it after the measurement window, and comes back with an actual number instead of a projected one.
That number changes two things at once: it lets a site get sized correctly today, and it lets that same site get sized for planned growth without guessing at the margin. Growth is where the nameplate approach compounds its own error: a system already oversized against real load gets a further growth margin stacked on top of it, and that second layer of padding is rarely questioned once the first one has already been accepted as normal.
Before You Size the Next One
If your last system got sized off a spec sheet or a previous installation rather than a measured load, it is worth finding out how far off that estimate actually was. Evoltix’s Total Cost of Ownership Studio can model what a right-sized system looks like against your own numbers, or an application specialist can walk through what a load study would look like for your site. Call +1 (855) 964-9274 or request a session.
Related Reading
An Oversized Backup Generator Costs You Twice
Telecom Tower Generator Runtime: How to Cut Fuel and Maintenance Costs