Use of Terms "Forward Acting" and "Reverse Acting" in PID Loops

[RBPLC]

Can someone provide clarification for me concerning the following:
I'm attempting to use a PID loop for a heating application. As I understand it, common terminology for PID controllers concerning Forward (Direct) and Reverse (Indirect) Acting is that a forward acting controller's output increases as the process variable increases (cooling) and a reverse acting controller's output decreases as the process variable increases (heating). From the documentation in the DMD PID help file, the terms Forward and Reverse acting are opposite of what I understand to be common nomenclature for PID controls. In DMD it appears that Forward acting means that as the output is increased the PV will also be increased (heating) and Reverse acting means that as the output is increased the PV will be decreased (cooling). It is easier for me to think in terms of the way Forward and Reverse are defined in DMD, but I believe that this is opposite of what is standard nomenclature for process controllers. Am I missing something?

[franji1]

If you have a heater, the control is Forward Acting - as you increase the Output, the temperature goes up.

If you have a chiller, the PID control is Reverse Acting - as you increase the Output, the temperature goes down.

[Controls Guy]

IMO, the industry standard is the same as DmD, PV moves with CV in forward acting.

[RBPLC]

I asked this question because my experience is opposite. I've dealt with process controllers like older Micristar and Watlow controllers and they use forward for cooling and reverse for heating. I see that in the Productivity it is used in the same manner as DMD.

[Controls Guy]

The one I can't stand is Allen-Bradley.  E=SP-PV vs. E=PV-SP.   I was griping to AB about it during a tech support call once vs forward/reverse and the tech chuckled and agreed.  I guess their way does avoid the issue we're seeing today, though!   

Quick check shows Love controls and Omron define it the way you're used to.  I've done a lot of temp control in the last 35+ years, and I guess more with PLCs than controllers, but I've used plenty of controllers, too.  I can't imagine how I got through all that time without realizing this definition was as common as it seems to be.  Weird.

[RBPLC]

It appears as though it might not be standard across all sectors and will be specific to a particular controller. I looked at the Click at it is defined the same as DMD.

[franji1]

Looks like this has been discussed a long time.  Found this post in PLCs.net from 13 years ago, talking about the confusion between A/B and Koyo (250):

http://www.plctalk.net/qanda/showpost.php?s=5db5298910d19db3e8d90e64e2cf8437&p=167056&postcount=4

I know Do-more based its terminology/behavior on DirectLOGIC (e.g. 250).

[RBPLC]

Worry more about how the system will act, than whether or not it is called a direct-acting or reverse-acting loop.

I agree with the above sentiment. I guess depending on your past experience and different control hardware the meaning of reverse and forward acting will mean different things. I guess the takeaway from this is that for every controller you'll have to read the documentation to see how it is defined for that specific controller.

[Controls Guy]

Looks like this has been discussed a long time.  Found this post in PLCs.net from 13 years ago, talking about the confusion between A/B and Koyo (250):

I don't get that thread at all.  I'm unaware of AB ever using the direct/reverse acting terminology, only the E=SP-PV and E=PV-SP.  Here are a couple screenshots of the PID configuration dialogs from Logix 5 and 500.  The last time I did a PID in 5K, it was the same (in fact, it was a 5K PID that I was calling tech support about the time I mentioned).