Page 355 - Sustainable On-Site CHP Systems Design, Construction, and Operations

328 Ca s e S t u d y 1

reporting, and data acquisition functions for plant operations. Control rooms include
multiple operator workstations, and an economic dispatch workstation. The cogenera-
tion control room also includes continuous emissions monitoring and an electrical
distribution/synchronizing panel. All plant control systems are fully backed up by a
UPS (uninterrupted power supply) and diesel generator.

Instrumentation
Extensive instrumentation is installed throughout the energy plant. Plant personnel
continuously monitor key process parameters to optimize economic performance. The
same database is used by the plant economic dispatch system. Historical data is then
collected and used to document fuel and water use and energy delivery, and to project
future energy needs. This database has become an invaluable resource for campus mas-
ter planning, engineering decisions, and individual system designs.

Real-Time Economic Dispatch
In August 2003, commercial electric purchases in New Jersey were deregulated. Prior to
that, Princeton purchased power at fixed day, night, weekend, and seasonal rates. Since
deregulation, Princeton has purchased power at a continuously varying wholesale
market rate. At night, prices are often as low as $20 per megawatthour—far below
Princeton’s marginal cost to generate power, and on a hot summer day the price for
electric power has risen to $1000 per megawatthour. Both liquid and gas fuel prices
have risen and have become more volatile since 2003. This provides strong incentive for
Princeton to be a very market-sensitive energy customer.
With the original tariff, the cogeneration system was run to follow campus load.
Any campus load not met by cogeneration was imported from the grid. To take advan-
tage of today’s energy market, Princeton plant operators need to regularly make changes
in power generation, fuel selection, thermal storage/discharge, demand-side manage-
ment, and the use of steam or electric-driven chilling.
In response to the wholesale market, a real-time economic dispatch system was
developed by Princeton and Icetec that continuously predicts campus energy demands
and market prices and then recommends the most cost-effective combination of equip-
ments to meet those requirements. The model inputs include real-time data for weather,
NYMEX gas and oil prices, campus energy demands, equipment efficiencies, and avail-
ability. By using this system the plant operator’s focus shifts from simply meeting
demand to delivering energy in the most cost-effective manner.
Princeton has found that in a highly volatile market, the cogeneration system oper-
ates fewer hours, but is actually worth more since there are more opportunities to shut
down cogeneration and purchase power from the grid less expensively, and more
opportunities to run at high load and avoid the highest-cost purchased power. The key
to Princeton’s economic dispatch is predicting those opportunities in advance and
being prepared to take advantage of them.
While this system could be fully automated, Princeton chooses to have plant opera-
tors use it as expert guidance—since there are times when safety, reliability, or critical
campus events are more important than short-term economics. The operators’ union
contract includes opportunities for annual bonus pay based on high compliance with
the economic dispatch signals. This has been a very successful program for both the
university and operations personnel.