Smart Building Systems for Adaptation to the Coming Smart Grid

Thomas M. Lawrence, #1 Richard T. Watson,*2 Marie-C. Boudreau,*3 Kyle Johnsen, #4

Javad Mohammadpour, #5 Lan Ding&6
#
College of Engineering, University of Georgia
Athens, Georgia, USA
1
[email protected]
4
[email protected]
5
[email protected]
*
Dept. Of Management Information Systems, Terry College of Business, University of Georgia
Athens, Georgia USA
2
[email protected]
3
[email protected]
&
Faculty of Engineering, University of Wollongong
Wollongong, NSW Australia
6
[email protected]

Abstract
The application of building automation and controls systems is now mainstream
across the globe, but there has not been a significant push toward developing tools
that will allow these new “smart” buildings to interact smoothly with the coming
smart grid. Electric utilities have been working for a while to develop tools and
procedures for what is needed on their end. Research is beginning on topics within
the built environment, but with disparate groups pursuing projects and ideas without
a grand unifying focus, and this unifying focus is increasingly important as the
spread of distributed generation of electricity through on-site renewable energy and
combined heat-power systems. Information gathering and modeling can reduce
energy consumption within individual buildings and the electrical grid, but a smart
grid will require new and more complex integrations of information gathering,
decision, and control. This paper discusses how the newly emerging field of Energy
Informatics can be used to help develop the tools needed for a smart grid. Current
building automation systems are not designed to collect inputs from energy market
signals, weather, and other appropriate data streams and make decisions based on
these data. One concept, termed Facilities Management and Modeling (FMM), has
been proposed by the authors, and this paper discusses in greater detail how this
can assist with developing for the smart grid/smart building interaction. This paper
also discusses the newly started investigation for the adoption of Energy Informatics
to a sample application on a major university campus located in the United States.

Keywords – Smart grid, energy informatics, smart buildings, modeling

 1. Introduction
We are just now entering a transition phase into the development and
implementation of smart electrical grids. The U.S. and Europe would likely
be considered the furthest along in this development, but interest in picking
up worldwide. Much of the early research and development work has been
done by the electrical utilities, for obvious reasons, but in reality all
consumers, generators, and transmission utilities will ultimately be affected.
The smart grid involves the generation and efficient use of electricity, with
much of this being consumed by the built environment. A system for
distributed energy generation from renewable and non-renewable sources
now exists and is growing in importance. The smart grid incorporates
aspects of electricity generation, distribution and consumption, and we see a
need for a unifying focus for research, policies, and standards associated
with the built environment in a smart grid context. Electric utilities have
been working for several years toward what is needed on their end. Research
continues on topics within the built environment, but with disparate groups
pursuing projects and ideas without a grand unifying focus. Such a unifying
focus is becoming increasingly important as the spread of distributed
generation of electricity through on-site renewable energy and as equipment
manufacturers begin to develop grid-ready devices [1]. Future smart grid
will see the need for communication and information sharing amongst all
entities – the “Grand Challenge” is developing and implementing the
infrastructure as well as the control functions to accomplish this.
Figure 1 illustrates the Grand Challenge that we envision as being
needed for the future smart grid. We see the need for developing the
technologies and procedures that will allow the connection between the
smart grid, the built environment, the “Internet of energy things” composed
of the literally thousands of sensitized objects that could exist in just one
building, and the overall information systems that analyze associated data
streams and control these attached devices.
A great deal of research and investment has already gone into studying
how a smart grid should be structured and developed, and information
systems analysis and analytics will be central to whatever form the grid
eventually takes [2]. Less has been done from the energy consuming side,
and looking at this from the built environment perspective is the focus of this
paper.
 Electricity generation,
and the “smart-grid”

New (dashed line):

Prior Power , information exchanged -
(solid line): sensitized objects, smart grid and
Power information system

INFORMATION
On-site renewable
SYSTEM Human
interaction and energy, combined
control heat/cooling/power
systems
Sensor and flow
BUILT ENVIRONMENT
networks
Energy using devices:
“The Internet of Energy Things”
Sensitized objects

Fig. 1 The grand challenge: Creating an information intensive and energy efficient built
environment using Energy Informatics

2. State-of-the-Art in Building Control and Automation Systems

The current capabilities of building automation and control systems is
advanced and is capable of doing a good job of the basic functions, such as
maintaining temperature setpoints or adjusting lighting levels according to
schedule or daylight availability. Suffice it to say that all major building
control and automation system equipment manufacturers are working to
enhance their capabilities for demand response management and interacting
with a smart grid. Many will state that their systems are capable of that now,
and they are to some limited degree, and private discussions with key people
at some of these firms confirms that there is more hype than reality in the
market at this time. Missing now are the fully developed and accepted
methods that will allow for ‘smart’ demand management and even
‘predictive’ demand management. Some of the tools and potential is being
identified. For example, Holmberg and Bushby [3] relay a background on
how BACnet, a widely used communications protocol for BAS, can help
integrate a building into the smart grid.
The format, content, and sources of energy management data within
buildings streams will also need further standardization and extension for
adoption beyond the building into demand-side management programs and
the smart grid [4], but current building automation systems are not designed
to collect inputs from energy market signals [5] and make decisions based on
these data. There is also the need for a larger scale system to allow a future
building to fully integrate into the coming smart grid. This need has been
echoed in one industry report [6].
New building and energy codes are coming into use that require some
electrical demand reduction measures, for example in both ASHRAE
Standard 189.1-2011 [7] and the International Green Construction Code
(IGCC) [8] that was released in March 2012. Standard 189.1 states that
“building projects shall contain automatic systems, such as demand limiting
or load shifting, that are capable of reducing electric peak demand of the
building by not less than 10% of the projected peak demand”. Similarly, the
IGCC requires that a new building project establish an open and
interoperable automated demand response infrastructure for key systems, and
that this demand response procedure is required to be able to reduce building
peak cooling or heating electrical demand by not less than 10% when
signaled by the electric utility or regional electrical transmission system
operator. Thus, there is a growing need for buildings to become more
intelligent so they can interact smoothly with the coming smart grid.
3. An Introduction to the Concepts of Energy Informatics and
Facility Management and Modeling
Energy and information are the central pillars of current and past
civilizations, and it is unlikely that their preeminence will dwindle in future
societies. Thus, it is surprising that this critical connection has not emerged
as a focus of joint research until quite recently. Energy Informatics is defined
as “analyzing, designing, and implementing systems to increase the
efficiency of energy demand and supply systems” [9]. The basic premise of
Energy Informatics contends that information has the power to reduce energy
consumption, which, in a simple fashion, is asserted as:

Energy Information + Optimal Controls = Less Energy

While Energy Informatics thinking has been in place for some time, its
formal emergence is a result of the development of mass scale, low cost
sensing devices. We now have the capability to collect massive amounts of
data with high granularity (i.e. high frequency and detail) to enable effective
application of optimization techniques and to create realistic simulations of
complex energy distribution and consumption systems. Analysis of these
various data streams provides an opportunity to recognize avenues for energy
reduction and modification of building operation to reduce energy. Thus, we
see Energy Informatics as having the potential to change the way that
buildings are design and operated. Energy Informatics can be a valuable tool
for the integration of building management systems and renewable energy
systems (on-site or in more remote locations on the utility scale).
Energy Informatics is, by nature, interdisciplinary in that it involves
analytical sciences, technical disciplines such as engineering, and in some
cases will involve public policy and economics. As applied to building
systems and their interaction with a smart grid, the technical areas of focus
will include building control systems, building energy modeling and
predictions, energy consuming equipment such as HVAC systems, electrical
utility integration and operation, and in some cases the interaction with on-
site renewable energy systems such as a photovoltaic array. Specifically,
when dealing with the interaction of buildings and their systems with a smart
grid, the domain of interest goes up several layers ranging from individual
pieces of equipment up through the regional or even national electrical grid.
This is illustrated in Figure 2.
Individual pieces of equipment:
1. Fault detection and diagnostics Equipment Level
2. Controlled by building systems

Optimize building energy consumption based on

actual and anticipated demands, balanced with
providing needed services

Bldg. A Bldg. B Bldg. C Bldg. D Bldg. E

....

Building Level
On-site renewable Optimized operation of
Central energy supplier
energy, combined energy supply systems
system (electrical grid,
heat/cooling/power based on actual and
chilled water, steam)
systems anticipated demands

Campus or Utility Level

Electricity generation,
and the “smart-grid”

Fig. 2 Varying scales for potential Energy Informatics applications

Lawrence, et al., [10] introduced the concept of “Facility Management

and Modeling” (FMM) for applications to buildings and the built
environment. FMM is a higher level of control, analysis and decision
making than what is operating at the Building Automation System (BAS)
level. As conceived, FMM is a method to enhance the process and remedy
some of the shortcomings of Building Information Modeling (BIM) and
BAS as they currently exist. We see this as the means to introduce the
means that will analyze the current and expected near-term future conditions
and make decisions as to what should the overall control strategy for the
coming hours or day should be that will optimize the building’s energy
consumption in relation to the needs or demands from the smart grid.
Current products available do not provide all the ‘smarts’ needed for
successful integration with the smart grid. We do not see FMM as being
another new product or piece of equipment that has to be added into the
building; rather it can be thought of as an evolution and enhancement to the
functions of existing BAS product offerings.
There are three core principles that guide the design of an FMM:
measurement; modeling and management. Measurement is important, as
without precise and frequent measurement, management becomes a guess
rather than an optimal decision. In the case of a facility such as a building
connected to the smart grid, this translates to creating a measurement system
that uniquely identifies every resource source and use location. Modeling is
a tool for understanding the consequences of circumstances and decisions
related to those circumstances. A building is typically subject to varying
conditions during the day as occupancy varies, resource loads change,
resource prices fluctuate, and weather conditions alter. Thus, modeling helps
to develop a prediction method for answering “what-if” type questions about
building operation that can be used to make smart decisions in the context of
a smart grid. Management involves the selection of an operational strategy
for that coming period of hours or days that will or may determine key issues
such as: equipment operational schedules and setpoint adjustments; building
temperature setpoints; ventilation and lighting control strategies; etc.
FMM requires multiple data streams to support the management and
modeling, and typical major ones are shown in Figure 3.
FMM produces two forms of output. First, it can signal actions to
multiple BASs to change the state of a building or campus of buildings.
Second, it can take direct action on those elements under its control. For
example, it might be able to adjust external blinds to reduce the sunshine
impinging on windows, or turn off certain devices during peak demand. As
well, it would produce managerial reports to identify high energy consuming
appliances, rooms, or buildings and recommend overall operational and
control strategies for consideration.
Every building and its corresponding systems is a somewhat unique
situation, in terms of operational aspects, thermal performance, equipment
limitations, etc. We see the ideal future smart control product(s) as being
those that will ‘know’ how that particular building will respond to different
operational setpoints or conditions. This knowledge can be derived from
observations of past operating performance, predicted occupancy levels,
weather, utility price signals, etc. Control system manufacturers are working
toward these goals, but still have a way to go to reach this optimal future.
There is also the problem of trying to implement smart grid functions
into existing buildings that have ‘not so smart’ controls. This is a major
impediment in the widespread implementation of smart grid concepts and is
a topic recommended for more study.
Major Building
equipment design Climate data

Building
design
BIM BAS
BIM BAS

Non-BAS
data streams
Building BAS
energy data streams
measurement
system Control
signals
FMM

Appliance
database

Weather Resources cost

data stream
data streams
Fig. 3 System interactions for determining strategies for smart grid
interaction using the concept of FMM

4. Example Applications
One example application of the integration of smart buildings with a
developing smart grid is the Pecan Street Project in Austin, Texas USA [11].
This project is built upon a municipal airport that was closed in 1999. The
redevelopment plans for this 700 acre (283 hectare) site were to create a mix-
used, sustainable urban neighborhood. This project received a grant from the
U.S. Department of Energy and other sources to create a living laboratory to
test concepts for integrating the built environment into a smart grid. Over a
five year timeline, concepts such as distributed renewable energy generation,
energy storage technologies, electric vehicle charging, smart electricity
meters and electricity pricing models will be tested and monitored. As of
now, this testing is approximately half way through completion. The Pecan
Street Project is involving roughly 1,000 private residences and 75
commercial businesses.
The development of smart building concepts involves a somewhat
different approach for residential as compared to commercial buildings, and
it is likely that these areas will require a different research and development
approach. The economic and public policy structure differs widely between
these two sectors as well.
The operation of campuses of buildings, such as in a university campus,
a large scale industrial or commercial building complex, or even a large
military base often will be different that that taken for a stand-alone building.
A campus of buildings often may have central energy systems, such as a
central steam or chilled water plant. These situations differ from other cases
where a group of buildings are using energy from an energy distribution
system in that a campus will have one common owner and likely a
centralized energy control structure. We see campuses as having a prime
potential for a coordinated electricity demand response and smart grid
integration.
The authors have initiated a test demonstration program on a central
chilled water production plant at the University of Georgia campus. The
initial concept for this demonstration was outlined in Lawrence, et al. [10],
and is intended to demonstrate the application of Energy Informatics
concepts to reducing peak electrical demand. This chilled water plant is
relatively new (startup occurred in 2011) and currently serves five buildings
with two chillers and is planned for future expansion to service 16 buildings
(existing and new) with the addition of six more chillers. This campus is on
a real-time pricing rate plan with the local electrical utility, and thus there are
direct financial benefits for the university that could occur with a successful
development and implementation of an electrical demand response program.
An illustration of the potential for smart operation of this chilled water plant
might have benefits in both operating cost savings as well as the positive
benefit to the grid of demand reduction can be seen in the sample plot in
Figure 4 of data of three summer days.
For example, note that on the weekday (Friday), the chillers were
running fairly hard even up until around 22:00 in the evening. The real-time
price of electricity remained high until around 20:00. The buildings served
by this system include a mix of classroom, meeting room and office spaces,
and it is very likely that an overall ‘smart’ integrated control concept for the
buildings and the chilled water system could make adjustments to zone and
chilled water temperature setpoints, etc. that would result in energy cost
savings. The cooling load is very much time of day dependent, with no real
peak demand reduction measures in place.
The general plan for this demonstration project is to collect data during
the 2012 cooling season then analyze the data through modeling and other
methods to create a means for predicting the operation of the chilled water
system and buildings connected to it. These would then be used to develop a
set of potential demand response strategies that will be tested during the
2013 cooling season. The ultimate goal is to develop a set of demand
response measures that can be implemented on a more or less automated
method for this system in the future.
1,000 Chiller Compressor Power, kW 40
left axis Outdoor Air Temperature, C
900
35

800
30
700

25
600

500 20

400
Approximate Real‐time Price  15
($ cents/kWh), right axis
300
10
200

5
100
Friday Saturday Sunday
0 0
10
12
14
16
18
20
22

10
12
14
16
18
20
22

10
12
14
16
18
20
22
0
2
4
6
8

0
2
4
6
8

0
2
4
6
8
Hour of Day

Fig. 4 Example potential application for Energy Informatics to a central

chilled water plant

References
[1] SmartGridNews.com, Smart grid trend alert: Smart appliances will soon proliferate. Oct 12,
2010. http://www.smartgridnews.com/artman/publish/Technologies_Metering/Smart-grid-
trend-alert-Smart-appliances-will-soon-proliferate-3138.html Accessed 25 August 2012.
[2] D.G. Holmberg, S.T. Bushby. BACnet and the smart grid. ASHRAE Journal, 51(11) (2009)
B8-B12.
[3] M. Smith, Utility Analytics: The Road from Data to an Intelligence Revolution, Utility
Analytics Institute (2012).
http://www.energycentral.com/reference/whitepapers/contact_vendor/103411/Utility-
Analytics-The-Road-from-Data-to-an-Intelligence-
Revolution?utm_source=2012_08_07&utm_medium=Email&utm_campaign=UAWeek&utm_
term=GRIDCOM
[4] J. Granderson, M.A. Piette, G. Ghatikar, Building energy information systems: user case
studies, Energy Efficiency Journal, June 16, 2010,
http://dx.doi.org/10.1007/s12053-010-9084-5.
[5] L. Wolffe, Higher Performance Buildings, Engineered Systems, 28 (10) (2011) 41-45.
[6] ASHRAE, Control of Demand to Drive Energy Efficiency in Coming Decade, ASHRAE
Journal, 53 (7) (2011) 8.
[7] ASHRAE, Standard 189.1-2011, Standard for the Design of High-Performance Green
Buildings. Atlanta: American Society of Heating, Refrigerating and Air-Conditioning
Engineers, Inc. (2011)
[8] International Code Council, Inc., International Green Construction Code.
ISBN 978-1-60983-059-5 www.iccsafe.org (2012)
[9] R.T. Watson, M.-C. Boudreau, A. J. W. Chen, Information Systems and environmentally
sustainable development: Energy Informatics and new directions for the IS community, MIS
Quarterly 34 (1) (2010) 23-38.
[10] T.M. Lawrence, R.T. Watson, M-C Boudreau, K. Johnsen, J. Perry, L. Ding. A New
Paradigm for the Design and Management of Building Systems, Energy and Buildings 51
(2012) 56-63. http://dx.doi.org/10.1016/j.enbuild.2012.04.016
[11] Pecan Street, Inc. “Smart Grid Demonstration Project at Mueller”,
“http://www.pecanstreet.org/projects/mueller/ , Accessed 25 August 2012.