5.1.9.2.3 are the electrical codes that apply to the 2 master alarm panels

A computer system substituting for a master panel must comply with 5.1.9.3 electrical and programming codes instead of 5.1.9.2.3, however, this code only applies if the computer system is substituting for the 24-hr panel.

There appears to be a gap in the electrical and programming codes that apply to a computer system that substitutes for the engineering master panel.

The code numbers have shifted by one number, but it's the same language since it was first allowed in 2005.  The computer used to substitute for the 24-hr panel must meet 5.1.9.4.1 & 2, but I don't see codes that apply to the communication between the initiating devices and a computer system that substitutes for the engineering master panel.

NFPA 99, 2005 5.1.9.4 Computer systems used as substitute master alarms as required by 5.1.9.2.1(2) shall have the mechanical and electrical characteristics described in 5.1.9.4.1 and the programming characteristics described in 5.1.9.4.2.

5.1.9.2.1 The master alarm system shall consist of two or more alarm panels located in at least two separate locations, as follows:

(2) In order to assure continuous surveillance of the medical gas and vacuum systems while the facility is in operation, the second master alarm panel shall be located in an area of continuous observation (e.g., the telephone switchboard, security office, or other continuously staffed location).

I’m not asking any questions – just pointing out an ambiguous part of the code.  If I’m wrong, well … that happens all the time.

1) A computer can substitute for only one of the panels, not both.  The code says “one” whether that is intentional or not.

Reference:  NFPA 99, 2018, 5.1.9.2.2 A centralized computer system shall be permitted to be substituted for one of the master alarms required in 5.1.9.2.1 if the computer system complies with 5.1.9.3.

5.1.9.2.3.1 codes apply to the communication between a traditional master alarms and initiating devices.  5.1.9.3 applies to the communication between a computer (used to substitute for a master alarm panel) and the initiating devices.  However, 5.1.9.3 only references the 24-hr location and does not reference  5.1.9.2.1(1), the engineering location.  The code only requires that 5.1.9.2.1(2) comply with 5.1.9.3.1 & 2.  I don’t see any requirements for communication between a computer in engineering and the initiating devices.  

Reference:  NFPA 99, 2018

5.1.9.3 Master Alarms by Computer Systems. Computer systems used as substitute master alarms as required by 5.1.9.2.1(2) shall have the mechanical and electrical characteristics described in 5.1.9.3.1 and the programming characteristics described in 5.1.9.3.2.

Well, folk, I dunno if I can add any clarity, but let me see.  Sorry if I just stir the mud even worse... 

The intent is that 2 wired panels be the "gold standard" against which other methods of communication are measured.  So, first, if you substitute a panel with a computer, the computer must act in every important respect like the panel it replaces, as detailed in 5.1.9.3:

Steve raises a valid point - 5.1.9.2.2 points one to 5.1.9.3.1 in which are rules for connecting the computer to the sensors.  Thus one might be tempted to assume that 5.1.9.2.3 .1 through .8 on communication don't apply to the computer-as-panel,  but I do not take that view at all - I believe they fully apply, and that the rules in 5.1.9.3.1 are additional, to be applied only the case of a computer-as-panel. This is because what you have to do to connect the computer to the sensors can be convoluted - the tricks manufacturers play to connect two alarm signals on separate panels with separate power supplies to a single sensor or switch may not work for a computer.  Worse, there are funky complications that can appear, some of which could be quite serious and hard to identify. 

The additional rules may seem murky, and yes Kevin they are intentionally ambiguous. The intent is to leave a little wiggle room to make it all possible. 

I can say that was the intent because we had a major logjam in the committee over this, and I wrote the original draft on which 5.1.9.3 was based. 

But the language probably can be improved, and you've heard the lecture on making submissions to NFPA ...