PLCS.net - Interactive Q & A

PLCS.net - Interactive Q & A (http://www.plctalk.net/qanda/index.php)
-   LIVE PLC Questions And Answers (http://www.plctalk.net/qanda/forumdisplay.php?f=2)
-   -   Temperature Controls in a Glycol Chilling System (http://www.plctalk.net/qanda/showthread.php?t=113456)

sparkie December 20th, 2017 12:25 AM

Temperature Controls in a Glycol Chilling System
 
I'm starting on the controls of a set of glycol cooling vats this week. Something the customer wants is a centralized HMI to control and monitor all the vats, but he also wants to be able to set the temperature at the each vat. Each vat also has a roughly foot and a half deep threaded thermal well for a temperature sensor. My questions are two-fold.

I'm thinking of using a Type J RTD to monitor temperature. Is there any reason these will not work for this type of application?

I'm running into an issue with changing the temperature at the vats and at the HMI. I have to parallel them so that neither takes precedence in sending the data to the PLC. I was thinking of using a temperature controller to read the RTD and send a 4-20mA back out to the PLC, but the issue still remains with viewing and changing the setpoint from each individual station. Any suggestions on this?

cjd1965 December 20th, 2017 12:58 AM

Hi why not use 2 hmi screens that reference same tag and have the PID control in the PLC?

Most temperature controllers can have a remote setpoint function wherw 4 to 20mA sets the setpoint or even use comms like RS485 etc

I would strongly consider the multi hmi route though even if one is full blown and the local one is cheaper but factor in a PID faceplate with a trend as a mimimum standard

willxfmr December 20th, 2017 02:16 AM

My first question would be what HMI software are you using FactoryTalk SE would make this pretty simple. Well... As easy as FactoryTalk makes anything else. But I agree with cjd1965. An HMI in a central location for monitoring, and as many as need to handle the local changes at the vats would be my first choice. Each HMI would be used to display information from, or give information to the PLC. All of the control programming should be handled in the PLC.

A type J thermocouple should work fine. If you want to bring the signal back as 4-20 you could use an RTD with a converter built into the unit. There may well be such a device for the thermocouple as well, but I'm not aware of one.

Bubba.

Mispeld December 20th, 2017 08:11 AM

If there is a PLC already involved, and that PLC has PID and trending capability, you will have significant advantages to using the PLC for control. Both for the lower complexity of set point and display management as suggested by other posts, as well as readily available tools for tuning the control loops. It is likely the PLC has a better documented PID implementation than a low-cost off-the-shelf single loop controller (SLC). On the other hand, a SLC may have an auto-tune feature not available (or an expensive add-on) in the PLC. Though any auto-tune capability must be used with care on an important control loop.

danw December 20th, 2017 08:41 AM

Quote:

Originally Posted by sparkie (Post 762840)
I'm thinking of using a Type J RTD to monitor temperature. Is there any reason these will not work for this type of application?

It will measure the temperature at the location of the thermowell, so the assumption is that the thermowell location is representative of the temperature of the entire bath.

A Type T thermocouple has better accuracy, out of the box, than a Type J and the Type T alloys are copper based, which are not subject to rusting, like the iron leg in a Type J.

A 3 wire RTD has better inherent accuracy than a thermocouple, using copper wire for connection between the element and the controller (instead of thermocouple extension wire), but you do have to pay attention to which wire goes where when connecting it to an input. RTD's do cost slightly more than thermocouples.

Temperature elements used in thermowells should be 'spring-loaded', so that the element tip is forced against the thermowell by spring pressure. And it pays to use some thermal conductive grease to aid thermal conduction. You measure the bore depth to get the right length spring-loaded element.

Mispeld December 20th, 2017 11:24 AM

+1 to danw post.

Unless you have a large number of thermocouples, you will probably be better off with signal conditioners (TC mV to 4- 20 ma) at the device instead of long runs of extension wire with potential wiring errors (e.g., uncompensated terminals, inadvertent cold junctions) going to thermocouple inputs on the PLC or SLC.

For cold temperature measurement we always use RTDs for accuracy and stability.

sparkie December 25th, 2017 06:42 PM

Thanks for the info guys. I'll be going with spring-loaded RTD's and I'm going to use panelmeters for the temperature display and signal conditioning. I'm going to try to talk him out of temperature adjustments at each station, as the control panel for temps is less than 50 feet away.

Saffa December 25th, 2017 09:01 PM

I don't see the issue with multiple HMIs? They can all reference the same setpoint without conflict.

sparkie December 26th, 2017 01:48 AM

Quote:

Originally Posted by Saffa (Post 763226)
I don't see the issue with multiple HMIs? They can all reference the same setpoint without conflict.

The issue is cost. 14 small HMI's would add up to quite a bit. When I asked what all this equipment needed to do, I was me with "I want it to make my coffee for me." Of course, if money is no problem then a coffee making function is no problem either :shoot:.

Saffa December 26th, 2017 05:51 AM

Ok, 14 is a bit excessive!

I think red lion's HMIs have a web interface version... chuck in a Wifi access point and give them any ipad as their mobile HMI.

I think some of the new Magelis panels can do the same but we don't ever make use of this feature so not sure.

Pete.S. December 26th, 2017 07:36 AM

I'd go for small HMIs for sure. There are small simple HMIs that are in the price range $100 to $200.

14 x $100 is $1400 so it's not a huge investment I think.

Aardwizz December 26th, 2017 09:01 AM

Are you aware of the existence of Single Loop Controllers ("SLC" -- not to be confused with the AB PLC with the same initials) ?

These cheap gizmos (~$100) do all the PID control for you, and have a display where the setpoints can be entered and temperature monitored. They also have communication capability (typically RS485 modbus) to talk to/from PLCs.

Your PLC would thus not need ANY AI/RTD or AO cards to talk to the thermocouples and glycol control valves, just a Prosoft module to talk to these guys. I think this is the kind of solution that your customer is expecting from you -- local control at the vat's; supervisory control at the HMI.

Technically / in theory you don't even need a PLC -- an open-source HMI like the AdvanceHMI that's popular on this site could do the Modbus interfacing with the SLCs. I've seen this sort of thing done; whether it's a good idea or not depends on one's ability to service/troubleshoot it down the road.

Anyway, just throwing it out there. Generally speaking, telling a client that he shouldn't want what he wants (local control) isn't a great way to get that client to come back to you. It's easier to keep clients than find new ones.

Good luck!

Pete.S. December 26th, 2017 09:31 AM

Quote:

Originally Posted by Aardwizz (Post 763248)
Are you aware of the existence of Single Loop Controllers ("SLC" -- not to be confused with the AB PLC with the same initials) ?

These cheap gizmos (~$100) do all the PID control for you, and have a display where the setpoints can be entered and temperature monitored. They also have communication capability (typically RS485 modbus) to talk to/from PLCs.

Your PLC would thus not need ANY AI/RTD or AO cards to talk to the thermocouples and glycol control valves, just a Prosoft module to talk to these guys. I think this is the kind of solution that your customer is expecting from you -- local control at the vat's; supervisory control at the HMI.

Technically / in theory you don't even need a PLC -- an open-source HMI like the AdvanceHMI that's popular on this site could do the Modbus interfacing with the SLCs. I've seen this sort of thing done; whether it's a good idea or not depends on one's ability to service/troubleshoot it down the road.

Anyway, just throwing it out there. Generally speaking, telling a client that he shouldn't want what he wants (local control) isn't a great way to get that client to come back to you. It's easier to keep clients than find new ones.

Good luck!

If all the client want to do locally is to change the setpoint then I think your solution is the best.

If modbus is fully implemented in the loop controller you could read and write all setpoints, process values, PID tuning/settings, operation mode etc from the PLC/HMI. That also means you can do full trends on the HMI if you read the SP/PV/OP trifecta on a regular basis.

Also good from a cost and installation point of view since you would wire up everything to the loop controller so there is a minimum of wires to run.

sparkie December 27th, 2017 05:24 PM

Thanks for all the input. I was actually playing with advancedHMI and the Logix simulators, but getting it to talk to them is quite a pain, so I wasn't really able to mess with it too much. Plus, with all of the stuff he wants to do with data I think a paid package would be preferable.

As far as using PID controllers for each individual VAT, it is a nice option indeed, especially since I'm going to use a PID controller for temp display anyway, and as mentioned before, I don't even really need the PLC. All of the logic can be done in the SCADA for cycle times and such, however there will also be glycol chiller monitoring and integration as well, plus I'm so sure that he will find other things he wants to add as time goes on I would rather just do the logic in the PLC.

It is excellent advice to refrain from "telling a customer what he needs," and it is okay to make suggestions based on your experience as to what may work well for a customer. I think the fact that the longest distance between a vat and the control panel is less than 50 feet, and that the customer will also have mobile access to the HMI on his cell phone is a good reason to refrain from having local controls, but once again, if it is something that he really wants it is something I'm more than willing to do.

Lots of design decisions to be made here, and I'm grateful for all of the input.

shooter January 4th, 2018 02:31 PM

I control this type of temperatures in holds of ships, with pretty simple Single Loop controls, they have everything inside one unit, complete with display, and are pretty cheap, They can have a communication like modbus or even etehrnet


All times are GMT -5. The time now is 05:47 PM.

.