Guides / Operating gaps
You came here for a number. Any number you find is somebody else's. The rate depends on your machine, your overhead and how you run. The method for getting to it, though, is much the same everywhere. So build your own rate instead of borrowing one that doesn't fit.
Ask what wire EDM costs per hour and you'll get answers ranging widely enough to be useless, because the people answering are describing their own operation. An operation with a paid-off machine running unattended overnight and one with a new machine and an operator standing at it have genuinely different numbers, and both are right about themselves.
A figure from somebody else won't apply to you. Here's how the work gets costed and what belongs in the rate. Build your own number and you'll be able to defend it, which matters more than just having one.
Wire EDM usually isn't quoted by the hour. It gets quoted on how much material the wire has to cut through, and time follows from that. The usual shape:
Length of cut, multiplied by the depth of cut, multiplied by your rate.
Length of cut is the path the wire travels, meaning the perimeter of the shape you're cutting, not the size of the part. Depth of cut is the material thickness, or the stack height if you're cutting more than one plate at once. Multiply those and you have the area of cut surface in square inches, which is the thing the machine is actually working through. Your rate is a dollar figure per square inch that you derive once from your own costs and then use everywhere.
This is why two parts of identical size can price very differently. A simple rectangular blank and an intricate profile of the same outside dimensions have very different path lengths, and the intricate one has far more cut surface even though it weighs the same and occupies the same footprint on the table.
Skim passes. This is the big one. The rough cut is one trip around the profile. If the finish or the accuracy calls for skim passes, the wire goes round again, and again. Two skims means roughly three times the path length, and if your estimate used the single-pass area, you have just quoted a third of the work. Whether you need skims is driven by the tolerance and finish on the print, so this is a question to settle when reading the drawing, not at the machine.
Everything that isn't cutting. Setting up and indicating the part, threading, and the handling on parts that need starter holes drilled first. On a job with many separate cut-outs, threading time alone can rival the cutting. None of that shows up in a square-inch calculation, so it has to be added on top rather than assumed into the rate.
To build the rate itself, the components are:
Divide the total by the hours, and convert to a per-square-inch figure using what your machine actually achieves in your typical material and thickness. That last part is worth measuring rather than taking off a specification sheet.
The distinctive economics of this process are that the machine can run without somebody watching it. That means the ratio of unattended hours to attended hours is what really moves your cost per part, more than shaving the rate. If you get long unattended runs overnight, your costs are genuinely lower than somebody doing everything on shift. You can't close that gap in a spreadsheet.
That's why a borrowed number is dangerous. If you take a rate from an operation that runs lights-out and apply it while running attended, you'll win the work and lose money on it, and the estimate will look fine the whole time.
None of this is worth much if the actuals never come back to the estimator. The rate you work out today is a guess until you compare it against what jobs really took. That loop is the subject of why job profitability numbers aren't trusted, and it applies here with particular force, because the inputs are measurable and the feedback is usually missing.
I spent a career in manufacturing, including a decade running a machine shop, so this is the method as it gets used rather than as it gets described in a brochure. The number itself is deliberately not on this page, because handing you somebody else's rate would be worse than useless. What I build is the part that makes your own numbers trustworthy: capturing what jobs really take, and getting that back to whoever writes the next quote.
See selected work →Plenty do, and it isn't wrong. The reason cut area is common is that it scales with the thing driving the time, so it's easier to estimate consistently from a drawing before anybody has run the part. Whichever you use, it has to reconcile against actuals afterwards or it drifts.
Directly, since thickness is one of the two terms in the calculation. What is less obvious is that very thick or very thin work can also change the achievable cutting speed and the number of passes, so the relationship isn't perfectly linear at the extremes. Worth measuring in the range you actually work in.
Yes, and this is where operations most often leave money behind. Setup, programming and proving out are close to a fixed cost on a new part and near zero on a repeat. Spreading them the same way both times means one job is subsidizing the other, and you can't see which.
I build tools that do the reading, extracting and checking around a quote so the estimator gets their hours back for the estimating itself. The rate and the judgment stay yours. There's more on that in how to quote machining jobs.
Related reading. The wider version of this is what actually belongs in a machining quote, and the feedback loop that keeps any rate honest is covered in why job profitability numbers aren't trusted.
First call's free. About 30 minutes, a straight conversation about how your operation really runs, not a demo. If there's something worth building, I'll say so; if there isn't, I'll say that too.