PLC and burner control regulations


Thread Starter

Anthony C.

I am a technician for industrial combustion systems and controls. I am having a hard time finding information about the current NFPA regulations concerning PLC's and burner controls. I am familiar with NFPA 8501/02. Is there any
clear-cut information out there to help me understand what is legal, or at least accepted by IRI, FM, etc.. PLEASE HELP
Responding to Dom's Thu, Mar 14, 12:01 pm comment:

You need the newer 85 series of the NFPA regs. They have incorporated more restrictive measures. If PLC's are involved, then, hardwired circuitry is also required. The implication is quite clear (at least to me)!

Phil Corso, PE
(Boca Raton, FL)
FM has there own set of specs as well, and they are often even more restrictive than NFPA. If your site is insured by FM or someone who follows their specs you should get a copy of those specs as well.
I dont use PLCs for burner control, i thought it was illigal. The idea being that on plc's things can be forced on i.e. safety mechanisms. i use the new honeywell control system and its ok

Bob Peterson

I have used PLCs for burner control. HOWEVER, the safetys have always been hardwired. Basically the PLC controls the temperature by modulating the gas valve, starts and stops the blower(s), etc., and monitors the safetys. It also starts and stops the burner as required.

The hardwired safetys always have the capability to override whatever the PLC is doing though, and shutoff the main gas valve in case of a problem (in the simplest case).

Bob Peterson
A bit off-topic, but how does NFPA (or similar)regulations for burner control safety cover the unlikely case of flame detectors failing in the ON state during normal operation, while the controller still thinks the flame is OK and therefore maintains gas supply (with the corresponding explosion hazard)? Is that case uncovered for the lower-range burners or for relatively non-dangerous applications? (I would
suppose the larger burners or critical applications would require flame detection redundancy, but I don't know)

Regards, Bob
Responding to Bob Gantt's Mon, May 6, 5:01 pm, comments:

Unfortunately your perception that a flame detector failure "ON" is unlikely is certainly not supported by statistics. I investigated several cases involving flame-detector failure. Although most were catastrophes involving only plant assets and topography, one killed 6 people.

Such fail-to-danger events must be detected if you are to meet NFPA 85 that NO single failure shall cause an unsafe condition. The single-failure criterion is easily accomplished through logic.

For those of you old enough to remember it has been my proven thesis that fault-mode diagnostics are more effective than voting systems when it comes to SIS performance evaluation. Such a fail-to-danger event as well as a fail-to-safe (an oxy-moron) one is covered in my paper "Probabilistic Risk Assessment of Safety Systems".

If you are unable to obtain meaningful info in the A-List Archives, I will be more than happy to provide additional detail.

Phil Corso, PE
(Boca Raton, FL - [email protected])