Page 104 - Zero Net Energy Case Study Buildings-Volume 2

CASE STUDY NO. 10
WEST BERKELEY BRANCH LIBRARY
Renewable On-Site Energy Supply
The renewable energy system as installed consists of 120 solar photovoltaic panels rated at 435 watts (DC) per panel, for a system size of 52.2 kW, with an estimated annual energy supply of 75,000 kWh. The final bid specification called for 160 panels rated at 315 watts per panel, for a system size of 50.4 kW. This was still an over-sizing of the system compared to the modeled building demand of 55,000 kWh, which would require only a 40 kW system. The oversizing by about 30% correlates directly with the measured overproduction of the system. (See Energy Production versus Energy Use: Zero Net Energy, below.) A certain amount of oversizing was deliberate, as noted in the ZNE Design Development Report to the client:
It is expected that higher performing panels will be available for comparable price at the time of bidding in late 2011, so that an excess capacity is likely. The excess capacity is desired since actual energy use may be higher than modeled. The models assume a cautious but nevertheless ideal schedule of operation and use, as well as many assumptions about plug load, the largest component of electrical load expected.
There are also sixteen solar thermal panels mounted on the north edge of the roof to avoid cast- ing a large shadow on the PV panels or skylights. The solar thermal panels provide the base hot water used in the radiant slab heating system. Originally specified to be tilted at an angle of 45°, the installed solar thermal panel array has a 15° tilt, matching that of the solar PV panels. This field adjustment essentially requires the heat pump backup system to operate more often, but was done to avoid casting long shadows into windows of neighboring apartments.
Energy Design Analysis and Energy Performance:
Modeling versus Post-Occupancy Measurements
Energy Use — Modeling
Energy modeling was carried out during the design phases using an early version of Design- Builder (v.2.1). The principal uses for the modeling results were to benchmark the energy de- mand and to confirm that a ZNE performance for the building would be feasible with the maxi- mum size solar PV system that could be installed on the roof. As design decisions were made, often based on cost criteria, the energy model was used to confirm that the energy demand did not drift upward.
DesignBuilder utilizes EnergyPlus as the simulation engine and, as a result, does not model well the passive ventilation in this particular building design. Some manual estimates were made for the energy savings from the use of the passive ventilation for daytime cooling and night purging, then applied to the modeling results. The model was subsequently updated using DesignBuilder v4.5.0.178.
The final design model yielded a total annual energy use of 47,700 kWh, or an EUI of 17.5 kBtu/ sq. ft. per year.
Energy Use — Actual Measurement and Comparison to Modeling Results
The original bid documents required that each electrical panel be equipped with branch circuit power meters for monitoring individual circuit loads. This specification was later changed to one power meter per panel, with each of three panels designated to all lighting, all plug load circuits and all mechanical equipment. The data from these three power meters was recorded over the course of a year (2014); the accompanying charts summarize the annual and monthly energy demand for each of these three principal types of electric loads.
92
Zero Net Energy Case Study Buildings: Volume 2