Software-as-a-service optimal scheduling of New Mexico buildings

Panel: 5. Energy use in buildings: projects, technologies and innovation

This is a peer-reviewed paper.

Authors:
C.Birk Jones, Mechanical Engineering University of New Mexico, USA
Chris Marnay, China Energy Group, Lawrence Berkeley National Laboratory, USA
Matthew Robinson, Mechanical Engineering
University of New Mexico, USA
Hans Barsun, Mechanical Engineering
University of New Mexico, USA
Leila Ghanbari, Mechanical Engineering
University of New Mexico, USA
Salman Mashayekh, Grid Integration Group
Lawrence Berkeley National Laboratory, USA
Wei Feng, China Energy Group
Lawrence Berkeley National Laboratory, USA
Andrea Mammoli, Mechanical Engineering
University of New Mexico, USA

Abstract

The UNM and LBNL are demonstrating the remote optimisation of a complex low energy building control using remote modeling in a software-as-a-service (SaaS) configuration. The building under this control is the UNM Mechanical Engineering (MechEng) building, which is a highly efficient building using about 40% as much energy as a typical structure of its size and function. Using the SaaS approach lowers energy consumption by a further third under favorable conditions, which happen in the shoulder seasons, and the energy bill is reduced by a similar amount. The building's HVAC system incorporates cooling assisted by a 232 m2 solar thermal array providing heat to a 70 kWthermal absorption chiller, as well as domestic hot water and wintertime heating. A 30 m3 hot thermal storage tank makes heat available at night for both heating and absorption cooling. Additionally, 350 m3 of chilled water storage shifts the considerable cooling electrical load of this high desert location off-peak. There are parasitic electric loads from a cooling tower pump, a cooling tower fan, and hot and chilled water circulation pumps, which become significant at very low energy consumption levels. The big energy and cost saving comes from more efficient use of storage. Depending on seasonal, weather, occupancy, and cost conditions, the cold storage tanks are only partially charged reducing losses. They are not used at all on weekends. A model been built on LBNL’s Distributed Resources Customer Adoption Model (DER-CAM) platform. A direct SQL interface between the building energy control system, and DER-CAM hosted on LBNL’s Berkeley server delivers daily scheduling based on weather forecasts, tariffs, etc. This paper reports results from tests completed in summer 2014, and prospects for upcoming installation at 3 other NM buildings with complex energy systems. New work has also been completed on thermal real-time monitoring of storage to achieve accurate control of state-of-charge, and low cost control using Raspberry Pie.