Carrying Out FAT at Site Instead of at Factory

A Factory Acceptance Test or FAT is done at the factory in case further work is required as a result of the test and to ensure the equipment is acceptable before it is shipped to the site.

That being said, usually it is the End User or Customer who determines when or where the FAT is conducted. If they will accept an on-site acceptance test in lieu of a FAT based on the vendors recommendations then that is up to all the parties involved.

FAT... great yet another acronym .... move over 'File Allocation Table' for 'Factory Acceptance Test'

It depends on what the contract says, including any 'Statement of Work' or tests described in the specification. Usually the customer has the right to have the FAT performed at whatever place he wants.

If the FAT is performed at a site the customer chooses (his place or an independent lab) the manufacturer has the right to witness the testing. The customer may then be responsible for the travel and expenses of anyone from the manufacturer's facility who needs to be there to witness the tests.

I'd recommend you look through the contract and all associated documents, plus any meeting notes between your companies to see what, if anything, is specified.

Good answers already but I'd like to add... ask yourself why the vendor wants to perform the FAT at your site rather than their factory. Understanding the value and intent of the FAT from the previous replies, what value/leverage/benefit does the vendor gain from the move:

• Reduced cost of the test (your utilities not theirs)
• Easier to test (associated processes are in-place)
• Increases likelihood to divert cause of issues to associated processes
• Increased likelihood of receiving acceptance even with problems due to the cost/time to uninstall and ship back to factory.

But, more importantly, ask yourself what additional risk does the end-user incur by moving the test. Hint, generally anything that provides benefit to the vendor is usually a net risk increase to the end-user.

I only ever agree to changing location of acceptance tests if there is a positive history with the supplier, and there is no 'new technology' involved, and I as the end-user representative can save the end-user time/money without increasing risk. And by 'new technology' I mean new to the vendor or end-user in the context of use.

ask yourself why the vendor wants to perform the FAT at your site rather than their factory.

That is not an usually request, especially if there are multiple components that are integrated together

Actually, it is not uncommon to do this on a large project. Several that I have worked on had 50+ cabinets 4'x4'x8'. That is a lot of real estate to cover.

The value to the customer, was being able to have most of their personnel get familiar with the equipment and programs. On one such job the company rented an abandoned J.C. Penny's store to stage all the equipment. As individual sections were programmed, they would bring the operators and E&I technicians from the plant (~1/2 mile away). The techs would step through simulating different I/O points, while the operators got familiar with navigating the graphics, and controlling the process. When you have 4-5 different E&I/Operator crews, that is a huge savings being across the street vs. 6 states away.

On a project that large, you get the spares along with the equipment, so if something doesn't work, you replace it, put a toe-tag on the bad part, and send them back as a group (with an RMA - so no cost to the customer). You would have to damage an inordinate amount of equipment for the large DCS/PLC vendors to even frown at you.

Of course... FAT's are not always viable... I've worked on quite a few projects in which FAT's were not feasible due to size/scope.

My reply was in the context of the OP's situation where it appears there was FAT originally scheduled at the vendor's shop but the vendor now wants to move it... assuming mid-project thus after post contract development.

Rhetorical of course, it sounds like this request is a change from an original agreement... why? When you answer the why and ID value/risk for the end user, the decision whether or not to accept the change in location for the acceptance test will likely make itself.

If such a thing is done at site, then by definition it cannot be a Factory Acceptance Test.

• If all else fails, ask the Project Manager. Which is what should have happened rather than asking a nebulous bunch at CR4, none of whom is liable for the consequences of any answer.

Call them "Site Acceptance Tests" (SAT)

Still remain open the issues mentioned by other contributors... why manufacturer doesn't wish to make the tests at his premises, and what implications will have this one to the customer.

Usually we are using both FAT and SAT with different purposes. What can be done on FAT remains for factory, and those test that are impracticable to be done on factory remains for SAT.

68

One possibility is that the manufacturer doesn't have the space for the equipment and is desperate to tackle other work to increase revenue. There is a risk to the project, that the equipment that arrives at site is not satisfactory for installation and has to go back to the factory for re-work, which is costly on project time, especially if the manufacturer has taken on more, and more lucrative, work that is getting in the way of rectifying the failed equipment.

So make sure the FAT is done properly at the F. It's the best place to remedy any defects, the equipment only travels once, and it arrives correct and fit for purpose.

We regularly had to perform acceptance tests on electrical control panels within our manufacturing premises followed by a further set of tests on site after the installation.

Tests done in your own premises and witnessed by the client confirm that the equipment of of the required standard before the client will issue permission the the equipment to be shipped to site. We always ran our own set of tests prior to the witnessed tests so that we were confident our kit would pass. In the UK there is a saying "Don't wash your dirty laundry in public" i.e. If you do have any problems, give yourself chance to fix them before the client knows about them.

There are some tests that can only be performed on site after installation. Earth bonding. continuity, crossed wiring, true compatibility with existing plant which does not always perform the way the specification claims or the way the client thinks it does. If there are subsequent site problems it helps you as the vendor to be able to point to your portion of the supply having passed FAT tests, and it helps the commissioning engineer to hone in on finding new problems because much of the kit can be eliminated as already tested and proved.

GA

We too do the pre-FAT testing without the customer, but must record EVERYTHING we find wrong and present that to the customer in a bound folder at FAT. Multiple pages of electrical drawings are always given to them with marks on them, because we put an incorrect number on a wire, missed numbering a wire, missed labels on electrical equipment, had incorrect colors on the drawings if a cable is not color coded as we thought it would be, and incorrect page number references on the drawings if we find them. Rarely do we find true operational problems, usually just the labeling / drawing issues, but we can't just change them for the customer FAT copy, we must record all of those changes for them. The idea is to show the customer we truly did FAT without them and this is the proof. And, of course, there are all the FAT documents signed and witnessed in the pre-FAT. A bit tedious, but it does show you inspected your product very closely.

If I was perfect I wouldn't be sitting here now. I would be on my 100ft yacht with a red head on my knee, a blond serving me drinks and a brunette waiting..... no it was the red head yesterday, I would have a blond on my knee with a brunette serving me drinks and a red head waiting in my bed. But non of us are perfect. When I started my first company it was as an electrical control panel builder. I had an agreement with my panel wire men. If they could find a mistake that I (and later on one of my designer / draughtsmen) had made, say on a drawing or a parts list, then I would buy the finder a pint of beer after work on Friday night. If I found any mistake that they had made, they had to buy me a pint on the Friday. The result was that they scrutinized all the work passed out onto the shop floor trying to score beer from me, and cross checked each others work before I went to test it to avoid having to buy me beer. It was the most effective and cheap quality control system I could have devised, and everyone was motivated to find mistakes with a friendly rivalry of trying to score points off each other, so very few mistakes got past and out to the customer.

Do you work in the perfect world? We are so pressed for time on all our projects, that we can't even have a review of our drawings by other engineers before building starts. We have to rely on the electricians to find the drawing errors. We have, several times, ordered the wrong parts because we must get parts coming before design is finished if we are to get done on time. The worst case was a panel box two standard sizes small by the time design was done. (it is a mess when the customer changes plans in the middle of design) Yes we found a way to work around that delay and get done on time - back planes are quick to obtain. As it is, I've worked for 6 weeks with just 3 days off to get a project done already. You can't find mistakes under conditions like that. If I tried that beer thing, I'd go broke, or end up in a hospital from extreme fatigue.

How large are your drawings sets? I'm talking three distinct sets of 50 to 70 pages at times, and a few weeks to draw them, with a 2 engineer staff.

Sounds to me that at the very least you should do the maths on comparing the cost of another engineer against the cost of the errors going through your shop. You also need to book that hospital bed now for when one of your guys gets sick or another job.

I was fortunate, I worked for five years as General Manager/Senior Engineer for another panel building company before setting up on my own. I went through the learning curve and made my mistakes with another guys money. At first we were only doing small panels but after a few years we were providing controls to automate whole process plants e.g. all the busbars and controls for a complete electro-tinning line, including 1700 drawings translated into Russian. We worked mostly in the chemical industry but also some food production lines and water utilities. I sold that company along with two spin off companies ten years after starting it. I then worked for the outfit I sold out to for fourteen months to integrate my 102 staff and my 'lean & nimble' working philosophy into their 1800 staff and top heavy bureaucracy. They had a quality control system that was all filling forms and ticking boxes. It didn't work because every body thought it a pain in the mule and treated it as a chore. It didn't generate quality but it sure generated paper.

No, it's not a perfect world...

Actually, on the small shop. The schedule is always tight. We done a set-up of a 1 million dollar process line, in our staging area in my shop. And as we were walking the customer through, I had the fitters and welders still working on it.

In the small design and fabrication shop, I've worked with engineers that were so anal, they won't release a drawing until every 'I' is dotted and 't' is crossed. Literally. They had weldments carried out to 4 decimal places, when 1/16" of an inch is called out.

So no, it's not a perfect world, but you do the best you can.

What costs of errors? The engineers are salaried, so when the schedule is tight and errors are encountered, if the hourly wiring techs can't get done in their 8 hour shift, guess who gets to stay and correct things? Works better that way, because the changes often involve drawing changes as well, so we just go record what we just changed. That's what we tell ourselves to justify 12 hour days and no weekends.

There are significant benefits to be achieved if the manufacturing process for both the supplier and customer are ready for this.

The majorpre-cursor is a "zero defects" process outcome based on quality assurance rather than quality control methodology.

Most GM and Ford (and other auto manufacturers) install engines that have been shipped from the manufacturers that have never been cranked over. They have reached a point of such high reliability of manufacture that testing is a cost burden on the industry that is no longer necessary. (Took a lot of effort from a lot of people, but is now expected practice.)

Is the supply chain ready for the consequences?

Yes it is a violation of a norm, but, the customer is ALWAYS right even when he is wrong. Remember that.

That said, we do 99% of our FAT here at the plant where the process skid, in our case, is built because if you find problems, it sure is a lot easier to fix where you have your tools and equipment handy, not to mention much quicker. Do you really want to transport tools to a site, or ship parts from your product back to the plant where it was constructed?

The entire purpose of a FAT is to find and fix problems before a SAT to avoid the issues mentioned above.

I certainly hope you are still negotiating terms prior to an order, not after. If you do FAT on site, have you negotiated the added costs involved in any potential repair so that the customer must cover these costs?

If you do the test at site instead of factory, it'll be called as (SAT) Site Acceptance Test. If both parties agrees for it, there should not be any problem if it works ok.