Page 404 - Sustainable On-Site CHP Systems Design, Construction, and Operations

Governmental Facility—Mission Critical 377

about whether precise terminology is necessary for practitioners who would under-
stand such distinctions in a specific application context.
But the distinction among load classes is significant. It is the main parameter for
matching electric to thermal load. In the most likely scenario, in which the EMA has
only enough funding to make incremental changes in a legacy DCOA, the separation
of load classes is necessary to keep capital and operational budgets honest. Consider
the following:
• The use of the word “emergency” in the NFPA universe of standards that cover
building safety is not coordinated with the use of the same word by the IEEE in
documents that deal with power systems at all voltage levels.
• The Federal Energy Regulatory Commission (FERC) uses the term “essential”
in its official rulings while the Joint Commission on the Accreditation of
Healthcare Organizations reserves that term for a subclass of loads in
hospitals.
• The National Electric Reliability Council uses the word “critical” in its Critical
Infrastructure Protection standard but the same word is reserved for a subclass
of loads in hospitals in Article 517 of the NEC.
• FERC refers to four classes of service to qualifying facilities: supplementary
power, interruptible power, maintenance power, and backup power.
• The term “mission critical” itself is copyrighted.

Without these distinctions it is possible for these technologies to fail to meet capac-
ity, reliability, or cost criterion.
Whether or not the CHP system supplies all or part of the DCOA or EMA facility
electric load, the COPS loads must be isolated from the rest of the facility’s noncritical
loads. The critical load isolation approach can be manual or automatic and can be
configured to incorporate dynamic prioritization of load matches to the CHP system
capacity.
In a peak-shaving or peak-sharing regime, the controls should include priority inter-
rupt logic that automatically suspends peak-shaving upon sensing a loss of adequate
power to the emergency loads. The same logic initiates retransfer or disconnect the peak-
ing shaving loads from the emergency or standby source to enable immediate transfer of
the emergency loads to the backup source. This reduces transfer time. Since the emergency
or standby power source is already running, the outage to the emergency loads is signifi-
cantly reduced.
The load to be tripped in a load-shed scheme should be large enough to compensate
for the maximum anticipated overload at one load-shed step. Choosing the number of
load-shed steps must be coordinated with the load and time required for each of the
systems shown in Table 23-1. When the core concepts of NEC Chapter 7 Special Systems
are placed side by side, it is easier to see the gap filled by Article 708. Within buildings,
these different power systems must be isolated from each other—typically by dedicated
switchgear, a separate conduit system, and possibly by a fire-resistant central chase that
ensures the integrity and survivability of power and control wiring.
More load than necessary may be disconnected for a less severe overload by this
strategy. On the other hand, it may result in a coordination problem among protective
relays if too many load-shed steps are involved. A typical load-shed strategy may only