IEEE Power & Energy Magazine - May/June 2018 - 49

records of load, weather, and solar data. historically, there were
often insufficient or incomplete solar data to perform forecasting and analysis of this caliber. however, with the recent advent
of comprehensive, high-quality meteorological data sets such
as Nrel's National solar radiation Database (NsrDB), the
electric power industry now has considerably more information
to better prepare for future integration challenges.
important considerations regarding Btm pv's prospective load reduction impacts include changes in the magnitude and timing of forecasted peak loads; changes in load
ramping and volatility; the frequency, timing, and duration of
light load conditions when load is low and Btm pv output
is high; and the overall impacts to typical load shapes. since
all of these considerations depend on the penetration of Btm
pvs, they require scenario-building and in-depth analysis of
resulting net loads (i.e., load minus Btm pv output).

Characteristics of BTM PV Fleets
Btm pv fleets include a large population of individual systems, perhaps totaling in the hundreds of thousands, installed
heterogeneously throughout a broad geographical region. for
many regions such as New england, measured performance
data are unavailable for much of the Btm pv fleet. metadata
concerning the exact design and location of all Btm pv systems are also often relatively limited. this lack of information is one of the challenges with respect to understanding
the seasonally and diurnally varying influence of Btm pvs
on load given a variety of weather conditions and as Btm pv
fleets become larger.
the realistic simulation of a decentralized fleet of smallscale, Btm pv systems should include all of the following:
✔ comprehensive, high-resolution weather data that includes the main drivers of pv performance
✔ some level of detailed information concerning the location and design of individual systems
✔ simulation tools capable of modeling pv system performance given a variety of weather conditions and a
diversity of individual system design characteristics.
in cases where detailed design characteristics of pv systems are unknown, it is possible to make informed assumptions
regarding these inputs and, ideally, to then validate the simulation results against measured data from installed Btm pv systems. We will discuss the development of such assumptions and
validation of results in the New england test case next.

The National Solar Radiation Database
Ground-based solar irradiance measurements needed for pv
modeling are sparsely available due to the expense and technical difficulty of installing and maintaining the requisite solar
radiation measurement network. to fill in this gap, the u.s.
Department of energy has funded the ongoing development
of the NsrDB by Nrel. in support of the Department of
energy's sunshot project goals, Nrel developed and released
the third generation of NsrDB, a comprehensive gridded
solar irradiance data set covering the continental united states
may/june 2018

and other parts of North, central, and south america. this
NsrDB release has significantly advanced the capability of
users to perform in-depth solar resource assessments within
the covered areas. Nrel, in collaboration with the university
of Wisconsin and the u.s. National oceanic and atmospheric
administration, made the most recent release of the NsrDB
available in late 2015. the data set covers the years 1998-2014
at a 30-min temporal resolution and is gridded at 4-km spatial
resolution. this means that there are almost 12,000 weather
grid points within New england.
Nrel used a physical approach to model surface radiation in detail by retrieving cloud and aerosol information
from Geostationary operational environmental satellite
imagery and subsequently used this information in a radiative transfer model. resulting NsrDB data includes the
three components of irradiance: direct normal, diffuse horizontal, and global horizontal. other ancillary weather data,
including dry-bulb and dew-point temperature, wind speed,
wind direction, relative humidity, and atmospheric pressure,
are downscaled from the National aeronautics and space
administration's modern era retrospective-analysis for
research and applications reanalysis data set.

BTM PV Fleet Simulation
the Btm pv fleet was simulated using Nrel's system advisor model (sam), a technoeconomic performance model used
as a desktop application and in the sam software development
kit (sDK). sam combines pv module and inverter submodels to calculate a pv system's performance given a weather

5+
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0

figure 6. A heat map showing the spatial distribution of
installed nameplate capacity (in MW AC) as of 31 December 2017.
ieee power & energy magazine

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - May/June 2018