Building Energy Performance

and Efficiency: Fact or Fiction?

Martin Fischer
Professor, Civil & Env. Eng.
http://www.stanford.edu/~fischer
fischer@stanford.edu

Director, Center for Integrated Facility

Engineering (CIFE)
http://cife.stanford.edu
Also affiliated with:
Precourt Energy Efficiency Center
Woods Institute for the Environment
Emmett Interdisciplinary Program in Environment
and Resources (E-IPER)
How we operate buildings today (with
scant historic data, last month’s bill) is
like driving a car by looking out the rear
window.

Real-time energy data allows us to drive

by looking at the side window.

A calibrated building energy

performance model allows us to drive
by looking ahead through the
windshield and gives us instruments.
2
 Goals for Seminar
• Discuss focus and approach of building
energy efficiency research at Stanford
• Find collaborators, prior work
– Establish cost and value-based metrics
– Close the loop on building energy performance at
the building, campus, regional scales
• Move towards “integrated operations” of
buildings and building portfolios
 Key Points
• Performance metrics are cost-based and not
uniformly established
– Is this the best we can do for energy efficiency
improvement in 2010?
• Modeling and monitoring energy performance is
difficult and current practice is greatly insufficient
– How can we improve strategically without validated and
useful benchmarks?
– How can we learn from one, many buildings over time?
– How can we claim to engineer and manage towards
high-performance buildings?
• We developed methods that help identify
performance issues
 Integrated building energy
management
• Across spatial scales:
Building – Building Portfolio – City
• Across the lifecycle:
Design predictions – operational measurements
• Across organizations
• Across data types:
3D building information models – many types of data – time
• Across utilities:
Energy – water – sewage – etc.
• For many performance goals (cost and value-based)
 Some current issues and questions
• Buildings are rarely built as designed
– Malmö: Non-tight buildings hinder sustainability
goals
• Buildings are rarely used as assumed by the
designers
– Occupancy, plug loads
• What to measure during building design,
construction, commissioning, operation
– Too much, too little data
– Methods to handle, make sense of data in building
life-cycle context
 Agenda
• Motivation for focus on building energy efficiency
• Cost vs. value-based metrics
• Comparing predicted vs. actual energy performance
• Managing building energy performance for a building
portfolio
• Managing building energy in the context of other
economic, environmental, and social goals
• Keeping multiple stakeholders focused on energy
performance throughout the facility lifecycle
• Discussion
Buildings use the most energy
in the U.S.

Building Energy Use:

39% of total U.S. energy

40% of carbon emissions
71% of electricity
54% of natural gas

Fastest growth rate!

Lawrence Berkeley National Laboratory

http://contexts.org/socimages/2009/06/16/the-effervesence-of-industrialization/comment-page-1/
Source:
Worldbank
 Energy Efficiency Offers Low or No-Cost
Carbon Reduction Options

Component by component analysis (e.g., “insulation”)

understates value of “whole-building” systems approach

11
Source: McKinsey Global Institute, 2007
 What does it
take to run a
theme park?

http://www.lasplash.com/publish/Domestic_150/
Disneyland_and_California_Adventure_Park_-
http://gocalifornia.about.com/cs/disneyland/l/bl_camap_dl.htm _It_s_Not_To_Late_To_Celebrate_.php
 Without electricity (and other forms of energy)
theme parks wouldn’t be nearly this much fun
• But … about half of the energy Disney uses goes to cooling
buildings
• And Disney expects to need a lot more energy in the future and
is working hard to reduce its energy consumption

Figure Courtesy, WDI

Do you want your entry fee to go
towards a chiller plant or a ride?

http://www.lasplash.com/publish/Domestic_150/
http://www.ameinfo.com/88546.html Disneyland_and_California_Adventure_Park_-
_It_s_Not_To_Late_To_Celebrate_.php
 On a m2 basis the “blue” building
performs poorly, i.e., it uses about 50%
more energy than the green building.

Figure Courtesy, WDI

 On a transaction per m2 basis the
“green” building performs poorly, i.e., it
uses about 80% more energy than the
green building.

Figure Courtesy, WDI

 Comparing Predicted vs. Actual Building
Energy Performance
Global Ecology Center
- Innovative HVAC system
- Data collection not reliable SF Federal Building
Y2E2
- Extensive measurements
- Our building (easy access)
- Innovative HVAC system
- Data collection started in
Dec. 2008
- Energy efficiency retrofit
project
- Controlled environment
- Typical HVAC system
- Data collection is starting in
February 2009
Main Jail North
- Natural ventilation
- Project completed
With:
Tobias Maile
Dr. Vladimir Bazjanac
Collaborations:
GEB
•Building Division, EETD,
LBNL
•Owners: GSA, Disney,
Senate Properties, Stanford
•A&E Firms: Arup, Granlund
•Systems Providers:
Honeywell, Siemens, UTC
17
Comparison example
• Barely meet code requirements
• Far away from performance goals
• More detailed data is needed
to pinpoint the causes

18
 Observed problem

1) HVAC systems do not operate efficiently

~ 10 – 40% wasted energy
(e.g., Haberl et al., 2003; Piette et al., 2001)

Waste due to inconsistent schedules and

control strategies, false operation,
malfunctioning components, etc.

2) HVAC design is not evaluated through operational performance

3) Today HVAC energy performance evaluation is
inconsistent, incomplete and time consuming
• Lack of methods to evaluate building performance (Xu et al., 2005)

19
 The big idea (2)
Space level System level Component level
P
R
relate E
D
I
C
T
E
D
aggregate Data from
Global
compare Ecology
Building

O
B
S
E
R
V
E
D

Value Intervention 20
 Each minute we are collecting
• 2,419 data points
• Sub-metering of light and plug loads for
each floor
• Four representative offices with very
detailed measurements

for the Y2E2 building

21
 Performance comparison methodology
Evaluate
HVAC
design
Energy model

Step 1: Preparation Step 2: Matching

Create, update and Update Predicted

adjust energy input data
model Predicted Step 3: Performance
Assumptions evaluation
Relate via
Establish product
product Detect Estimate
breakdown
Step 1:structure
Preparation breakdown differences impact
structure

Observed
Assumptions
Setup data Observed
collection data

Legend Improve
Performance building
Tasks
data operation

Feedback Assumptions Actual building 22

 Energy simulation input

HVAC system Weather

Simulation control
Internal loads parameters

15,000 input objects Results

- Valve position and flow rate correlate
- System works mostly during occupied hours

Work Work Work Work

week week week week

24
Night purge on the 1st and 2nd floor seems to be on a regular
schedule rather than dependent on outside and inside
temperatures

Should be and Should be closed

is open

close

open
25
 Progress – Y2E2
1. Identified about 50 instances of performance issues
2. We and Facility Operations have resolved/are resolving
performance issues (as appropriate)

Instances of identified issues Resolved

Problem categories issues
EPCM CEE243 Designer FacOps
Set point differences 3 3 None 1 2
Scaling issues 20 15 None None 20
Incorrect design assumptions 4 None 3 None -
Cycling problems 3 2 None None 3
Incorrect control strategy 16 13 None None 5
Sensor problems 6 5 1 None -
Time effort (approx. hours) 450 1,100 200 ? -
26
 Energy performance of Y2E2

Source:
Arup’s
Calibrated
Y2E2 model
report

• Actual operation is close to design baseline model (code)

• Actual operation is far away from original design goal
• Calibrated
11/2/2009 models show about 50% Tobias
improvement
Maile compared to calibrated baseline
27
Santa Clara County Jail Energy Efficiency Retrofit
1. Geometry model is complete
2. HVAC model in progress
(End of November)

28
Santa Clara County Jail Energy Efficiency Retrofit
3. Identified (30) new
sensors to be installed
(420 existing sensors)
4. Installation of those
new sensors is almost
complete (End of
October)

Example: Condenser Loop

Water flow rate and return
water temperature are
missing to calculate energy
supplied by condenser loop
(Energy = flowrate * const *
temperature difference)
29
Next steps for the Santa Clara County Jail
Project
• Finish EnergyPlus model to complete step 1 of
methodology (End of 2009)
• Use methodology (step 3) to identify performance
issues (End of February 2010)
• Recommend retrofit options based on problems
identified and other retrofit opportunities (End of April
2010)
• Track installation of energy retrofits to obtain an
accurate as-built model and track performance of the
building (through 2010/11)
• Communicate results to County (and Stanford and Rest
of the World) with conference on energy efficiency
retrofit process 30
4D benchmark: A method to quickly
identify underperforming buildings with respect
to energy and water consumption within a
building portfolio
Main Researcher:
Mauricio Toledo, Ph.D. Candidate

Support from:
Center for Integrated Facility Engineering
Stanford Housing
Universidad de Chile, Santiago
 Where to intervene to reduce energy
consumption?
Stanford SH 260 buildings
12,000 students
4.5 million sqf

How hard is to
achieve further
savings?

Toledo 2008
PhD Proposal
vs 32
 Intuition
ORGANIZE IDEF0 diagram
PRODUCT &
CONSUMPTION
NO PORTFOLIO DATA
APPROACH

Benchmark
energy & water
DISPERSE consumption
Identify
Assess
operational PROVIDE
METHOD &

INTEGRATE
underperforming
requirements
Identify
REPRESENTATION

underperforming
BUILDING
DATA buildings buildings
Display
consumption & NO FORMAL
underperforming METHOD
buildings
Inconsistent
PARTIAL TOO FEW OR NO Inaccurate
OR NO OPERATIONAL Time
PROVIDE CONTEXTBENCHMARKS REQUIREMENTSPROVIDE METHOD consuming
TO CONSUMPTION FOR HANDLING
DATA SEVERAL O.R.
Toledo 2008
PhD Proposal 33
 Method to manage energy efficiency
for building portfolios
method to generate
Model for
PBS & CBS
modeling of product formal representation
and consumption data
Building
portfolio Benchmark
Energy & water
energy & water
consumption consumption
Assess
Weights: wij operational
Requirements: Ri
requirements
Identify
underperforming
benchmarking & buildings
operational requirements Display 4D benchmark
model that identifies
consumption &
method underperforming
underperforming
buildings & displays
buildings consumption data

Method for
Method for Method for
operational
portfolio visualization of
requirement
benchmarking consumption data
checking

Toledo 2008
PhD Proposal 34
 Create 4D-CAD model

Bldg height
Bldg name (text)

3D_DO_SQF_Branner

3D_DO_STU_Branner

3D-CAD model:
3D_DO_SE2_Branner

3D_DO_ABS_Branner
- Campus map with streets layouts
3D_DO_SE1_Branner
- 3D buildings of interest (gray clustered layers)
- Other 3D campus buildings for reference (single
3D_DO_HIG_Branner

layer)
Naming convention

Toledo 2008
PhD Proposal 35
 4D Benchmark
conceptualization
Utility Type Temporal Aggregation

MO CU SE
Monthly Cumulative annualized Comparison to itself for
consumption consumption baseline year (2005-06)

ABS MO_ABS CU_ABS SE_ABS

Utility consumption in
absolute terms

STU MO_STU CU_STU SE_STU

Consumption
per student

SQF MO_SQF CU_SQF SE_SQF

Consumption per
square foot
3D-CAD model

HIG MO_HIG CU_HIG SE_HIG

Highlighting of bldgs
exceeding operational req.

Consumption Type SQF
SQF
SQF

STU
STU
STU

ABS
ABS
ABS

Toledo 2008
PhD Proposal 36
 Development Strategy Simulator
Martin Fischer
Collaborators: Ryan Orr, Jonghoon Kim, Tanmaya Kala, Matt Hart,
SU-QU Overseas Seminar Participants
Northern Doha,1973-2007
Sprawl dense high
So … do we have enough?
Who decides?
 On what basis?
Why do we have to work for the information?

© Stanford
Let’s make the information work for us.
Traffic Electricity Gas

Manufacturing Manufacturing

Electricity
Traffic

Gas
Commercial Commercial

Logistics Logistics Water

Jan 2010 Jan 2010 Jan 2010

Workers Manufacturing

Water
Commercial

Logistics

4D simulation Jan 2010

Workers

Jan 2010
Sewage

Manufacturing
Contribution

Sewage
To GDP Commercial

Logistics

Manufacturing Jan 2010

Commercial
GDP

Logistics
Cost of Cost of
Cement Legend
Infrastructure Building

Environmental metrics
Social metrics
Cement
Cost

Cost

Economic metrics
Jan 2010 Jan 2010 Jan 2010
© Stanford
4D model plus metrics
© Stanford

Compare Development Scenarios

• Product mix
• Timing
• Impact of codes, regulations
• Resilience of development strategy to changes in …
Scenario 1 Scenario 2
 Development Strategy Simulator
Input Output
Integrated
 QEZ master plan  4D visualization
(3D model, schedule) Transparent  Analysis for metrics
 Relevant data and (4D + data)
coefficients
Simulate
3D Model
Simple
strategies
(Analysis using GSM of building
floor area from 4D model,
seamless integration) Electric Water Gas
Use Use Use
Schedule
Water

Demand
Visual Jan 2010
Electricity
Gas

(4D and chart simulation)

© Stanford
Current Practice: Compiled from several industry case studies and surveys

Metric Assessment
# Stakeholder groups represented ~2

Objective definition Informal and incomplete

Number of options generated and analyzed ~3

Types of analyses performed ~4 (cost, program, area, aesthetics, structure)

Process efficiency (% Value Added Time) ~40%

Formation of Decision Rationale Informal and political

Process sharing and understanding informal and inefficient

Gane V. , and Haymaker, J. (2008). “Benchmarking Conceptual High-Rise Design Processes,” CIFE TR # 174

Clevenger, C., Haymaker, J., and Swamy, S. (2008). “The Importance Process: Enabling Creativity in Performance-based Design through Systematic, Model-
based search of Multidisciplinary Impacts,” World Sustainable Building (SB) Conference Proceedings, Melbourne, Australia, in press.

Haymaker, J Chachere, J, and Senescu, R (2008) "Measuring and Improving Rationale Clarity in a University Office Building Design Process” CIFE TR 178

Flager, F. and Haymaker, J. (2007). “A Comparison of Multidisciplinary Design, Analysis and Optimization Processes in the Building Construction and
Aerospace Industries,” 24th International Conference on Information Technology in Construction, I. Smith (ed.), pp. 625-630.

© John Haymaker, PhD, AIA, LEED ap 2009

Reid Senescu and John Haymaker
SOCIAL AND TECHNICAL INTEGRATION

PIDO PIP MACDADI

Integrated Concurrent Engineering session to

Automate, Communicate, and Decide
© John Haymaker, PhD, AIA, LEED ap 2009
Reid Senescu and John Haymaker
 Automatically Generate & Analyze Design Spaces
Total Lifecycle Operating Costs
($/ 30 years)

Height(
m)
Radius
(deg)
© John Haymaker, PhD, AIA, LEED ap 2009
Reid Senescu and John Haymaker
Enable Multi-Stakeholder Decisions

Benefit = Analysis x Preference

B
A

© John Haymaker, PhD, AIA, LEED ap 2009

Reid Senescu and John Haymaker
Laboratory Results

Metric Practice New Methods

# Stakeholder groups represented ~2 6

Objective definition Informal and Clear, more complete

incomplete
Number of options generated and ~3 hundreds
analyzed
Types of analyses performed ~4 (cost, program, area, ~ 7 (structure, cost, energy,
aesthetics, structure) lifecycle cost, aesthetics,
convenience…)

Process efficiency (% Value Added ~40% TBD

Time)
Formation of Decision Rationale Informal and political Clearer, fairer

Process sharing and informal and Formal, accessible

understanding inefficient
© John Haymaker, PhD, AIA, LEED ap 2009
Reid Senescu and John Haymaker
 TYPICAL HOUSE LOBLOLLY HOUSE
90 % ON SITE 70% OFF SITE
10 % OFF SITE 30% ON SITE

Conventional Parts vs. Integrated Component

Assemblies
CIFE Conference © KIERANTIMBERLAKE 2007
 Round trip distance traveled
from home to factory daily

PAUL 24 miles
TONY 24 miles
ERIC 44 miles
JOHN 26 miles

Materials And Resources – Loblolly Travel

Distance
CIFE Conference © KIERANTIMBERLAKE 2007
 Next steps
• Application of DSS, 4D Benchmark, Comparison Method
– More case studies
– Test impact on stakeholders and development processes
• Development of tools
– Add infrastructure elements
– Add ecosystem functions
• Interactions between natural and built environment
– Interconnections between issues
– Links to GIS, regional analysis
• Data
• Underlying science basis
• Vision
– Tool/method for democratic and efficient engagement of all key
stakeholders in building energy efficiency
So? …
• How can we establish building energy performance metrics
that are meaningful and stick?
• How can we establish methods to assess building energy
performance so that we can believe and compare the
assessments?
• How can we retool the management of the building
lifecycle (design, construction, commissioning, operations,
retrofit) to scale energy efficiency to every building on the
planet?
• Other important questions?
• Where can we collaborate most productively? Include
others?
55