Pathways for Intermittent Power: Integrating Cloud Forecasting with the Management of
Energy and Water Infrastructure to Accelerate the Adoption of Solar Power Generation
Institutions and Principal Investigators
University of Arizona:
Alex Cronin
Nathan Allen
Gautam Gowrisankaran
Mike Leuthold
Vincent Lonij
Stanley Reynolds
Young-Jun Son
Department of Physics
Department of Biosphere 2
Department of Economics
Department of Atmospheric Sciences
Department of Physics
Department of Economics
Department of Systems and Industrial Engineering
Los Alamos National Laboratories:
Feng Pan (Currently on his sabbatical in SIE at UA)
GOAL – Solar facilities produce intermittently with by far the highest production during clear, sunny
periods. We will explore new ways to compensate for the intermittent output of solar power plants through
grid-scale experimentation that combines dynamic control of existing water infrastructure, cloud
forecasting models, PV solar generation, and battery storage. The work we propose here will combine
expertise in Physics, Engineering, Economics and Atmospheric Sciences to address the intermittency
problem using the Biosphere 2 as a Model City.
RATIONALE – Solar power utilization at the utility scale is a Grand Challenge. The intermittency and
cyclic nature of solar energy is seen as among the biggest hurdles to their large-scale adoption.
Intermittency limits the adoption of solar power by utility companies and industry because they require
reliable, predictable power generation. Discovering new ways to compensate for intermittency will
accelerate the adoption of solar power, will reduce CO2 emissions, and will improve national security by
conserving nonrenewable resources.
PROPOSED IDEAS - Intermittency can be mitigated with energy storage or spinning reserves, but these
approaches alone are costly. This project will build on the preliminary work funded by WEES to develop a
solar-aware industrial-scale smart grid at Biosphere 2. This project will dynamically regulate energy use of
existing water infrastructure (water pumps, treatment) based on measured and predicted solar power
production. With the help of NSF-SEP funding we intend to expand the scope of the WEES research work
from preliminary data collection and modeling to grid scale experimentation. NSF funds will also enable us
to incorporate energy storage with the B2 smart-grid to smooth out fast transients caused by both solar
intermittency and the switching of industrial size water pumps. A modest amount of storage can increase
the applicability of our load control algorithms to a wider variety of loads.
The work we propose here will combine expertise in Physics, Engineering, Economics and Atmospheric
Sciences to address the problem of intermittency using existing water infrastructure and cloud forecasting
models to reduce the need for energy storage and minimize development costs. Our research will verify
how existing water infrastructure can be used with new control algorithms to help industries and utility
companies tolerate intermittent power sources such as photovoltaics (PV) and improve the economics of
renewable energy development. We will evaluate the cost effectiveness of our new methods by applying
existing economics models to new data. This will allow us to develop new estimates of the cost of solar
energy in the presence of a solar aware smart grid and energy storage.
Biosphere 2 can be operated off-grid as a power island with PV and diesel generators, or on-grid as an
industrial scale power consumer. Various sub-units at B2, such as the 4-acre indoor biological laboratory,
the 28 residential buildings, the visitor center, and 70kW of PV panels on-site, each have different power
consumption/production profiles. These power demands make B2 a model city, with energy profiles similar
to hospitals, residences, department stores and power plants. This makes B2 an ideal and unique location
for smart grid research that can be scaled-up and implemented by larger industries.
This project will focus on the following research questions:
1) Investigating the effect of solar intermittency on power quality (voltage and frequency stability) at the
Biosphere 2 micro-grid when it is operated on- and off-grid. We will measure time-series of power
production from the three existing PV arrays equaling 60kW at B2, with data acquisition at 1-second
intervals. These measurements will form a baseline for the load regulation scheme proposed below.
2) Estimating the value of advances to forecasting techniques in reducing the social cost of solar facilities.
Forecasting will be done using satellite data, sky cameras, and regional networks of irradiance sensors.
Advances include more sophisticated simulation models in addition to physical facilities.
3) Evaluating the optimal use of water pumping and battery storage as mechanisms to mitigate the effects
of solar power intermittency by storing energy. We will develop algorithms to make optimizing power,
water management, and batter storage decisions, given the available power demand and solar output
forecasts. Battery banks will be installed and used to test findings obtained from our models.
4) Understanding the economic effectiveness of improved forecasting, batteries, and water pumping on
reducing the true social welfare costs to the Model City of solar PV. We propose to estimate the cost
savings by combining our technologies with an optimizing, economic model of reserve operations, battery
storage and water pumping. This work will allow us to evaluate the extent to which the Model City should
invest in battery storage and optimal water pumping technologies as a complement to solar PV.
COLLABORATION – This project will leverage the existing collaboration between scientists from
Physics, Atmospheric Sciences, Economics, and Engineering departments at U of A. In addition we will
collaborate with Dr. Feng Pan at Los Alamos National Labs who brings grid-scale power modeling
capabilities to our team. The diversity of our team will greatly broaden the impact of our work.
TRAINING and OUTREACH - This NSF-SEP project will train students in physics, economics,
atmospheric sciences, and systems/industrial engineering to work together and to use Biosphere 2 as an
energy systems research laboratory. Biosphere 2 provides Science Education Tours for over 100,000 people
per year. The theme of these tours (human-environment interactions) includes energy research. Research
activities are incorporated into tours, so our NSF research methods and findings will be communicated
directly to the public and other UA researchers in this venue.
WHY THIS IS GOOD FOR U OF A AND FOR ARIZONA - Arizona is ideally suited for this study,
because of the huge potential to scale our proposed method; states like Arizona and California use 15% –
19% of their total energy consumption to pump, convey, treat, and distribute water. Experience we gain by
implementing a solar aware smart grid at Biosphere 2 on a 100 kW system can be scaled, first to a 10 MW
scale system like the city of Tucson Water utility and then to the GW-scale pumps at the Central Arizona
Project (CAP).