How much time does it to take to replace a PLC

sparkyinak

Member
Join Date
May 2016
Location
Ketchikan, AK
Posts
79
This is real generic question with too many variables like the multitudes of different PLC's. It's just for ghits and siggles, for I think for a interesting discussion

Let's say you have have this machine that makes widgets. It's not a complex process like petrochemical mixing or running a smelting plant but widgets are the companies bread and butter and the process is complex enough that warrants a reliable PLC. The manufacturer of the widget machine worked with you to develope the machine. The PLC required 40 or so I/O's which they listen to your recommendations for a PLC preference for whatever your reasoning is.

Years later the widget machine is still punching out widgets day in and day out. Suddenly the PLC gives up the ghost. Total lost failure and need replacement, like yesterday. You Are handed the job to replace it, and I need the widget maker back up and running like yesterday.

Not to make anyone or company look bad here, answer the following reasonable as possible from you are at in the industry, like plant electrician, general electrical contractor, specialty contractor, manufacturer rep, etc. This is just strictly out of chariousity as a sit here on the mend.

Questions to be answered:

1. What 40 or so I/O's reliable PLC did you recommend for the widget maker and why?
2. The PLC you specified, are replacement PLC are stock on site?
3. If no to 2, what is locally available or expressed freighted in from afar?
4. How much time would it take you to replace in kind the PLC from answer 1 to get the machine operationally ready? So the time here must include the time for everything from unwiring and removal of the bad PLC, wire labeling, notes, whatever to safely rewire the PLC, install new PLC, reterminate all wiring properly, testing a must(which was 100% correct, (good job), program, verify, etc. The clock stops when widget maker is release for production.
 
If there are only 40 discrete IO points involved, then the installation and wiring should be easily completed within one day. The complexity of the program would determine how long it would take to rewrite the logic. I am assuming that there is neither a backup nor a printout of the program?
 
The fastest way to do this type of job is to rip everything out and start from scratch, including agreeing on the functional description.

Actually the "Fastest" way to do this job, would be to install the exact same PLC that went bad, load an existing copy of the program back into it, slap it into run and go get some lunch. 🍺


Unfortunately this option is not always available.
 
Last edited:
Actually the "Fastest" way to do this job, would be to install the exact same PLC that went bad, load an existing copy of the program back into it, slap it into run and go get some lunch. 🍺


Unfortunately this option is not always available.

Yes, or just hit it with the magic wand :geek: and make it work!
Actually the question is TOO generous. Basically I would say that it will take some time. Or less. Or More. Or much more. You would have to develop also some documentation after all (I would make it before).
 
This is from a recent job. My customer's grinding machine had a PLC that was no longer available. When the PLC failed they were able to locate someone with a similar model which had been taken out of service. That user was willing to send the EPROM to my customer. While they were waiting for the EPROM I visited the site to gather information about a controls rebuild on the assumption that even if the EPROM got them back into production it was time to modernize. That was on 15 December.
A week later the EPROM did not get them going. I visited again and they decided they were going to modernize in two steps. The first step is to replace the PLC with something current to get them running again. They will then produce an inventory of parts to ride through a longer period of downtime during which they will do a complete mechanical and controls rebuild. That decision was made on 21 December.
I put together a bill of materials for a replacement PLC. At their request it was for Mitsubishi L-Series PLC. Customer ordered it. On 27 December I got the purchase order, the schematic for the machine, and a printed copy of the PLC program that supposedly was controlling it. The serial number of the machine on the prints I got was not the same as the serial number on the nameplate of the machine.
I marked up a copy of the electrical schematic to show how to move the existing wires to the new PLC and translated the ladder logic from the program they gave me. It was simple ladder logic with timers and counters so I had reasonable confidence that the translation would be close to operational. The only potential issue was that the original included a MCR instruction and I could not be sure that the implementation of the MCR zone was the same in both PLCs.
We started the changeover on Thursday, 5 January. Continued on Friday. I had a prior commitment on Monday, 9 January so I could not be at the site. On Tuesday, 10 January we turned the power back on. Due to my unfamiliarity with Mitsubishi it took the better part of the day to get the program loaded to the PLC.
It took another week to get the machine back into production. There were several discrepancies between the printed copy of the program that I received and how the machine actually needed to be run, including some signals that were present on the old PLC but didn't appear on the schematic I was given. There were also difference of opinion about how the machine was supposed to operate between the line supervisor, the maintenance technician and the engineering department. There were a couple of "oh, that's never worked right, can you fix it" issues.
Final acceptance came on Tuesday, 17 January.
The I/O count was similar to what you've cited. The new PLC has three 16-point input modules and three 16-point output modules. It was just over a month (including Christmas and New Year) between first contact and final acceptance. I was on site a total of seven days between the day we removed the old PLC until acceptance. All work was done on first shift. There was nobody available from the customer on any other shift.
 
Last edited:
There were also difference of opinion about how the machine was supposed to operate between the line supervisor, the maintenance technician and the engineering department. There were a couple of "oh, that's never worked right, can you fix it" issues.

this is the part that takes you the most time.

I did the same thing last month, it took me 2 weeks , because there was not an electrical diagram nor the ladder diagram for the plc.
 
I see two potential scenarios in your question, not sure which you meant:

1) What machines have you walked up to and replaced a PLC in, and what struggles did you face?
2) If I were designing the machine today, what would I choose so that I could replace the PLC quickly later on?

I think a couple people have answered scenario #1, so I'll choose scenario #2.

If you have a machine built with a Siemens S7-1200 (using a CPU plus a few IO modules) and the PLC dies, you could have the system up and running quickly. I'm assuming you've already done enough trouble shooting to know the power is still good and which modules need to be replaced. Whole thing could easily be done in under an hour, with minimal room for error.

The PLC program can be stored on a memory card. This can easily be swapped from PLC 1 to PLC 2. No download is likely required (as long as a memory card was included in the original PLC), unless whatever destroyed the PLC destroyed the memory card as well.

The wires don't need to be individually transferred, as the terminal strips snap on and off.

Honestly, I'd probably spend more time wrangling the thing on/off the rail than I would anything else, although at least minimal testing should be done to ensure nothing else is wrong and the terminals didn't get swapped.

Spares are hypothetically available at the local distributor, but if we're talking about the far future than that's harder to guess. You can hope you have spares on site, but it depends how many of the widgetmakers you have for that to be realistic. Siemens usually promises spares will be available for 10 years. Also, future models are typically backwards compatible/spare part compatible with older ones, so you may luck out and be able to buy new hardware that supports your old program.
 
Honestly, I'd probably spend more time wrangling the thing on/off the rail than I would anything else, although at least minimal testing should be done to ensure nothing else is wrong and the terminals didn't get swapped.

Honestly, in your setup, I would probably spend more time haggling with the customer on the price and doing the paperwork then on the work itself, don't you think? :D
 
I had one that I did a few years back... the PLC was a Lighthouse http://www.lighthouseplcs.com/ , I had the logic printed out and was able to install a 1400, I was able to swap the CPU and rewire the new one in about 2 hours, I went with a 1400 because of the physical size and layout was close to the original one and did not have to change wires, I think it took about 10hrs to figure out the logic so for a straight digital about 40 I/O total was about 12 hours plus startup debugs

The machine was a crate flipper / conveyor, the people would fill up crates about 4ft square with peaches in the fields and bring them to the plant, there they would place the crates on this machine that would slowly flip the crates and feed them onto a conveyor that would feed the wash/sorting machine

It was like this one

https://www.youtube.com/watch?v=Yymj57-WViw
 
If you have a machine built with a Siemens S7-1200 (using a CPU plus a few IO modules) and the PLC dies, you could have the system up and running quickly.

You could say the same thing about practically any PLC currently on the market. Of course, this assumes that the failure occurs in the very near future.
 
Having been there myself, I would do the following.

What plc's are in the plant and what plc reps are in the area?
what plc programming are you familiar with?

pick a plc that you are familiar with.
choose the plc as close to the dead plc size as you can.
look at the cad drawings (if any) and choose the I/o based on what you are looking at.
if no cad drawings, you MUST map the I/o yourself, be careful with the outputs, the machine will move.
you WILL have to draw a sketch of the new I/o wiring, very important.
on the left side of the paper is the inputs and wiring #'s, in the middle is the descriptions, on the right side is the new inputs.
do the same thing for the outputs.

if no program or documentation exists, meet with everyone and have a discussion about how the machine operates. those that run the machine as the best source. agree on the sequence and write the code.

if you are not familiar with plc programming, I would recommend that you get an integrator to do the job. Costly up front, but less downtime.

hope this helps,
james
 
I have this discussion quite a bit actually, involving all kinds of aging automation equipment. People make economic decisions at the point of purchase without a lot of regard to the future viability of the cheapest product, then make an add-on blunder by not buying the spares at the same time. Fast forward 5 years and the machine is down, the parts are no longer available or the current version has incompatible firmware and someone needs to rip and replace, which involves programming. Even if we can do it in 12 hours on a simple machine, as Geniusintraining showed (nice work by the way), replacing it with a spare of the original make from off the shelf in the plant would have been 2 hours or less. So what was the cost of that additional 10 hours of down time compared to the cost of keeping the spare on hand? It's almost never a consideration, which is tragic in my opinion.

Sometimes the bean counters have their heads up their rears on this issue. Rant complete...
 

Similar Topics

Hello. I have a customer developing a customized OPCUA client that needs to gather data from ControlLogix PLCs, but this is not a PLC integrator...
Replies
2
Views
1,104
Good Evening , I'm interested in the Bentley Promis.e Design Software. Have many of you used it , and how much does it cost ? Is there any...
Replies
0
Views
1,083
how much does cydec cost? Specifically the runtime. and is there a tag limit cost?
Replies
6
Views
3,496
I got a price on a AB PanelviewPlus 700 and it was $3,329.00. I am shocked ,I know I go thru Culture Shock everyday but Sticker Shock of this...
Replies
17
Views
4,281
I am new In a CCW and as a beginner I am trying to learn programming but i am noticing that my CCW software is taking around 1 minute to download...
Replies
2
Views
71
Back
Top Bottom