[Bldg-sim] RMESE Meeting March 19

[Peter Ellis]

Join us *Tuesday, March 19* for the next meeting of the Rocky Mountain
Energy Simulation Engineers (RMESE). Five masters and PHD candidates from
the University of Colorado will have 15 minutes each to present on their
respective energy modeling projects for their research. The
presenters/presentations for the evening will be as follows:

*Neal Kruis*

How do you account for heat loss from your building foundation? Many energy
modelers either drastically simplify this heat loss in their models or
neglect it entirely. This is because the multi-dimensional, large-timescale
nature of foundation heat losses can become very computationally intensive,
to the point where it becomes time prohibitive for most modeling projects.

My research aims at answering several unanswered questions about foundation
heat loss:

·         How significant is foundation heat loss relative to other loads
in the building?

·         How good are our current tools at estimating these losses?

·         What can be done to improve theses estimations and reduce
computation time?

·         What are best practices for insulating building foundations?

*Anna Osborne*

A grant awarded to the International Center for Appropriate and Sustainable
Technology (ICAST) from the U.S. Department of Housing and Urban
Development (HUD) is funding a study to assess the effect of energy
conservation measures (ECMs) and behavioral change measures (BCMs) on the
energy use patterns of 800 units of multi-family low-income residential
properties. Since the effects of ECMs are known and calculable, the main
goal of the study is to determine the degree of change that can be achieved
through education of the occupants and combined effects of ECMs and BCMs.
Anna has also developed a Matlab program which performs complex linear
regression on utility data to assess the degree of efficiency of the
building in question. The energy modeling component involves a sensitivity
study on occupant behavior by adjusting schedules and setpoints as well
as comparison to the Matlab program. The results are still preliminary.

* *

*Ben Brannon*

Ben Brannon will be presenting the initial development of his thesis on the
Modeling and Control Design and Field Experimentation of a Thermally
Activated Residence.  The project stems from a home in Missouri that
attempts to direct heat obtained from massy walls to beneficial locations
around the home and control radiative heat transfer.



*Lincoln Harmer*

This project serves as a proof-of-concept that calibrated building models
when applied real-time information provide enhanced energy management and
diagnostic capabilities. Rather than comparing against a static baseline as
is customary in Monitoring Based Comissioning (MBCx), the building models
used in this project are dynamically kept in tune with the actual building,
over time horizons stretching several time scales. A straightforward energy
management application involves determining the percent deviation in
facility natural gas consumption or electricity use between the actual
building and its model over the last day, week, month, and year; a
predefined threshold of permissible deviation (say +/- 10 ) would then
trigger an alarm or alert to the building management staff. Selecting the
right performance metrics will be part of this research. Envisioned
examples of extended diagnostic capabilities include a) fault detection and
diagnosis (FDD) such as finding the most likely HVAC system parameters that
explain the observed consumption patterns and thus detecting system
degradation, b) prediction of facility energy use and electrical demand for
the next day or week, and c) load aggregation for multiple buildings to be
served by an electrical supplier in a deregulated utility context. The
simulation models employed may be white box models based on first
principles, statistical black box models using only monitored data, or
inverse gray box models that combine reduced order building physics with
simplified models for the energy systems. These inverse gray box models
involving both first principles as well as parameter estimation appear as
likely contenders.  The expected result is a software demonstration of
using a building model and measured historical and/or real-time building
information for building performance evaluation. While many of the research
questions may be answered using surrogate data, we’re collecting real-time
data from two facilities and are aiming to prove the concept on both
buildings using a real-time field implementation.



*Ryan Tanner*

My research is entitled *Stochastic Optimal Control of Mixed Mode Buildings
Considering Occupant Driven Uncertainty*.

I use offline model predictive control to optimize the controls of
(simulated) mixed mode buildings, a process which results in 'optimal
control datasets'. Next, I use machine-learning algorithms to derive viable
BAS control rules from the optimal control datasets. EnergyPlus has been
the building energy modeling tool used throughout my research, providing
virtual test-buildings to try different controls in. Throughout all
simulations, occupant behavior is accounted for by one of two methods. (1)
Co-simulation using the building controls virtual test bed (BCVTB) to
couple the EnergyPlus simulation with occupant behavior algorithms written
in matlab. (2) Writing occupant behavior algorithms directly into
EnergyPlus via the Energy Management System (EMS).

This will be an exciting group of presentations, so make sure you join us
to support our future modelers!!!

We will be meeting at *The Tavern Downtown (1949 Market Street, Denver, CO
80202) at 6:00 pm Tuesday, March 19 *for* *the five short presentations and
drinks.

Please RSVP to Jessie Jones at jjones at rmhgroup.com or (303)312-4641 so that
we can make sure our reservation is large enough. This is not invite only-
so pass the word on to coworkers, colleagues, and other industry
professionals who may be interested.

Thank you,
*RMESE Steering Committee

*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/bldg-sim-onebuilding.org/attachments/20130311/05941638/attachment.htm>