Understanding Per Diem in Engineering Field Work
Per diem calculations in engineering aren't as straightforward as looking up a GSA rate and calling it a day. I've been doing field cost estimating for over a decade, and the thing that trips people up most is that the official government rates are just a floor, not a ceiling, and they don't actually cover what you're spending. The Strom Engineering Per Diem framework is essentially a standardized method for breaking down daily field costs into line items that hold up under audit. It comes out of the civil and structural engineering contracting world, where crews go out to sites for days or weeks at a time. The idea is to take the raw GSA per diem rate for a location and rebuild it with actual operational costs layered on top - not just meals and lodging, but equipment downtime, site access fees, meal allowances that actually reflect what food costs in that area, and the overhead of supporting a crew on the road. Most people I see trying to use this just copy the GSA table and slap a markup on it. That doesn't work. The system is designed so you start with the base rate, then adjust for factors like remote locations where everything costs more, or urban sites where lodging eats your margin. I built my first real model around 2018 when we were bidding a bridge inspection contract in rural Montana. The GSA rate for that area was laughably low - like $89 a day for everything. Our actual cost came out to about $240 per crew member per day when you factored in that the nearest affordable lodging was 45 minutes away and meals meant driving into town every single time. The gap between the published rate and reality is where projects either make money or bleed out.
Building the Calculation Model
Start with the GSA rate for your specific location and date range. You can pull these from the GSA website directly. Then you're going to break it into three buckets: lodging, meals and incidentals, and your engineering-specific add-ons. The lodging piece is usually the loudest variable. In high-cost cities like San Francisco or New York, the GSA rate covers maybe 60 percent of what an actual mid-range hotel runs you. In rural areas it's worse because the published rates are based on whatever the cheapest motel charges, not what you'd actually stay at when you need WiFi and a quiet room to do report work. For meals, the M&I rate is fixed nationally by GSA, but again it's a theoretical number. A crew of four engineers eating at diner prices in the Midwest will burn through that allowance in a couple days. I learned to budget meals at about 1.4 times the M&I rate for field crews and 1.2 times for solo engineers who can cook in their room. The add-ons are where most people cut corners and then wonder why their bids come in under budget and their projects lose money. You need to account for per diem days that aren't full work days. If a crew arrives on a Friday afternoon and leaves Monday morning, you're still paying for Thursday night's hotel and Friday's meal allowance even though they only worked half a day. I track this by counting arrival day as 0.5 and departure day as 0.5, then applying the full rate to any overnight in between. This usually adds 10 to 15 percent to your total per diem cost compared to just multiplying full days by the rate.
Then there's the equipment and material transport angle. When a crew is out of town, their normal vehicle may not be sufficient. You might need a cargo van for instruments, or a truck for surveying gear. That's a separate cost from per diem but it's directly tied to the number of per diem days. I include it in the same spreadsheet so it doesn't get missed during bidding. Fuel for that extra vehicle across a week-long job in remote territory can run $400 to $800 depending on terrain and distances between sites.
Get the Full Details
Common Mistakes That Cost Money
The biggest mistake I see is using a single per diem rate for an entire project when the crew visits multiple locations with different cost profiles. You'll be bidding a project that has a two-week stint in downtown Chicago followed by a month in western North Carolina. If you apply the Chicago rate to the whole thing, you're either overbidding the Carolina portion and losing the job or underbidding it and eating the difference. I separate each location into its own section of the model and calculate independently. Another thing people miss is the tax implication. Per diem paid above the GSA rate is generally considered taxable income to the employee. So if you're reimbursing someone $200 a day when the GSA rate is $140, that $60 difference shows up on their W-2. For a week-long assignment with a two-person crew, that's roughly $600 in additional tax withholding that you either absorb or factor into your rate. I usually build in a 20 percent tax buffer on any excess reimbursement, which covers federal and state without getting into the weeds of individual tax situations. Remote site logistics are the third trap. I worked a project in northern Alaska where the published per diem didn't account for the fact that the only available lodging was a company bunkhouse that cost $95 per bed per night but required a minimum 30-night commitment. The GSA rate for that area was $178 total per diem, which sounds fine until you realize it includes $53 for M&I and $125 for lodging, and the bunkhouse alone consumes almost all of that lodging portion. We ended up negotiating a group rate through the contractor that brought it down to $70 per bed, but getting that required advance contact and written agreement before mobilization. If you wait until the crew arrives, you're stuck with whatever's available.
Download and Template
I put together a spreadsheet that handles most of this automatically. You input the locations, dates, crew size, and it pulls the GSA rates, calculates the adjusted lodging and meal costs, applies the half-day rules, and sums it all up with the tax buffer. It's not fancy but it's been through about fifteen real bids and caught enough errors to be worth using. You can grab it from my site's resources page. The file is called strom_perdiem_model.xlsx and it's built for Excel 2016 or later. I also include a reference sheet with common adjustments for things like holiday surcharges, seasonal pricing spikes, and the remote site multiplier I mentioned. If you're doing this kind of work regularly, I'd suggest pairing the spreadsheet with a simple log where you record actual costs after each job. The difference between your estimates and your real spend is the data that makes the next bid better. Most people never do this and then wonder why their per diem numbers drift further from reality every year. A single project's actuals usually takes about twenty minutes to enter, and over a year that adds up to enough data to calibrate your assumptions without guessing. One final note on what this system doesn't handle well: short-term assignments under three days. The overhead of setting up lodging, getting oriented, and dealing with travel delays eats a disproportionate share of the per diem budget. For jobs under a week, I tend to round up the per diem by about 25 percent to account for the compression effect. It's not in the model automatically because it's more of a judgment call, but it's something I apply consistently now after losing money on a couple too many weekend-warrant jobs where the crew spent more time figuring out where the hotel was than actually working.
