Environmental Engineering Reference
In-Depth Information
total energy required to match the output of the renewable sources to loads
over the times of interest that may be as long as several years to account for
weather variations. Fluctuations on the order of minutes (Figure 1.1) might
be smoothed by the installation of batteries, supercapacitors, or flywheels.
Thermal energy storage may be useful for storing solar energy generated at
midday to meet evening demands.
High energy storage applications are currently limited to PHES and CAES.
The greenhouse gas output of an electrical system will be a function of the
resources running on the system and their respective emissions. A resource
on the margin or the next resource in the dispatch order will be the one
that powers the energy storage. The higher the penetration of renewable gen-
eration in a given system, the more likely the resource on the margin will
be carbon-free. Likewise, the greater the penetration of variable renewable
generation, the greater the need for firm capacity to respond to loads in the
absence of sun and wind.
An addendum to an EnerNex Corporation report on the electrical grid in
Colorado shows that at a 10% penetration of renewables without storage (a 324
MW pumped storage facility at Cabin Creek near Georgetown, Colorado), the
cost to integrate the renewable generation resources doubled.
5
Although the
Cabin Creek facility has proven cost effective and technically capable in wind
integration, a modern facility would be more effective by achieving faster
response time and greater ability to adjust pumping loads based on the genera-
tion available at any given time. A best case scenario combines the traditional
pumped storage with advanced variable speed technology. The EnerNex adden-
dum to a report titled “Wind Integration Study for Public Service of Colorado:
Detailed Analysis of 20% Wind Penetration”
6
shows a sensitivity analysis of
the value of pumped storage relative to the sizing of Cabin Creek. It states:
In addition to the analysis of wind forecasts, several sensitivity cases
were run using the same input data. The Company and TRC felt it was
important to include the sensitivities of pumped storage capacity on
wind integration costs. This was evaluated by simulating the PSCO
system with a varying number of pumped storage units. The existing
Cabin Creek units were used as templates: each unit has 163 MW gen-
eration capacity and 117 MW pumping capacity. The scenarios modeled
were: 0, 1, 2 (current capacity), 3, 4, and 6 units. Generation and pump-
ing capacity as well as the pond size were scaled for each scenario. The
integration costs reported here were created using an old (pre-WWSIS),
un-smoothed wind forecast (and, therefore, a high geographic diversity
of wind farms) and $5/MMBtu gas prices. (Table 1.1)
The results show a decrease in integration cost from about $10/MWh
with no pumped storage available to a low of about $3/MWh for 6 units
(3 times existing capacity). The implication from this sensitivity analysis
is that the ability of pump storage units to pump when there is excess
wind and deliver energy to meet the variability needs is of significant
value both from an integration cost standpoint and from an overall pro-
duction cost perspective.
