1

Automated Demand Response Strategies and Commissioning Commercial Building Controls

Mary Ann PietteLawrence Berkeley National Laboratory

mapiette@lbl.govdrrc.lbl.gov

California Commissioning CollaborativeJuly 10, 2006

Sponsored by U.S. DOE, California Energy Commission (Demand Response Research Center),

San Diego Gas and Electric Company,and Pacific Gas and Electric Company

2

Presentation Objectives

• Introduction to Demand Response• Building Design and Operations Framework• Building Commissioning and DR• Encourage Discussion linking DR and Cx• Collaboration Opportunities with the CCC

3

Objectives and Background• Objectives

• Develop DR communications infrastructure, demand response control strategies, and assessment of field tests.

• Evaluate role of commissioning in execution of strategies• Improve understanding linking demand response and

energy efficiency• Definition of Demand Response

• Short-term modifications in customer electric use in response to dynamic price or reliability information.

• Relevance• Electric systems more vulnerable to outages with age,

load factors decreasing, T&D and new capacity investments reduced, real time pricing promoted.

4

Building Design and Operations Framework

Efficiency andConservation

(Daily)

Peak LoadManagement

(Daily)

DemandResponse(Dynamic

Event Driven)

Motivation- Environmental

Protection- Utility Bill Savings

- TOU Savings- Peak Demand

Charge savings- Grid Protection

- Economic- Reliability

- Emergency- Grid Protection

Design- Efficient Shell,

Equipment & Systems

Low Power DesignDynamicControl

Capability*

Operations- Integrated System

Operations

Demand -Limiting and

Shifting

Demand -Limiting,

Shifting, orShedding

Initiation Local Local Remote

*Prefer closed loop strategies, granular control

5

Results on Automated-DR• Established capabilities of

current controls and communications with EMCS and XML

• Demonstrated initial design of signaling infrastructure and system capability

• Demonstrated large sheds can take place without complaints

• Demonstrated range of strategies to produce sheds and capabilities needed

• Average reduction 10% among 22 buildings, up to 50%

Aggregated Demand Saving, Sept 8th

0

1000

2000

3000

4000

5000

6000

7000

0:0

0

1:0

0

2:0

0

3:0

0

4:0

0

5:0

0

6:0

0

7:0

0

8:0

0

9:0

0

10

:00

11:0

0

12

:00

13

:00

14

:00

15

:00

16

:00

17

:00

18

:00

19

:00

20

:00

21

:00

22

:00

23

:00

De

ma

nd

[k

W]

Albertsons B of A (B) OFB Roche USCB Total Savings Baseline

0%

10%

20%

30%

40%

50%

60%

Ech

elon

UC

SB

OFB

50 D

ougl

as

AC

WD

2530

Arn

old

Gile

ad 3

42

Roc

he

US

PS

Gile

ad 3

57

Tar

get

Cal

EP

A

IKE

A

Cha

bot

Alb

erts

ons

Ora

cle

300C

Mal

l

B o

f A

NA

RA

CIS

CO

OS

Isof

t

450G

G

Ave

rage

Dem

and S

avin

g W

BP

%Site Average

Total Average10.3%

6

Automating Demand Response

1. PG&E defines price schedule

2. Price published on LBNL XML (eXtensible Markup Language) server

3. Clients request price from server every minute & send shed commands

4. EMCS carries out shed automatically

Internet andPrivate WANs

= Price Client= Pilot site= Price Server

= Development Site

Polling Client &IP-Relay Software

Internet& private WANsLBNL

Price Scheduler

1

2 3

Electric Loads

CC C

EMCSProtocol

3

Gateway

Price Server

Test SitesC = EMCS Controllers

Electric Loads

CC C

EMCS Protocol

C

IP-Relay

Infotility

PollingClient

4 4

2003 test was Gateway only2004 was Gateway or Relay2005 both

Akuacom & Infotility

7

Critical Peak Pricing(PG&E Example)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0:0

0

1:00

2:00

3:0

0

4:00

5:00

6:0

0

7:00

8:00

9:0

0

10:0

0

11:0

01

2:00

13:0

0

14:0

0

15:0

0

16:0

0

17:0

0

18:0

0

19:0

0

20:0

0

21:0

0

22:0

0

23:0

00:

00

Ele

ctri

city

Pri

ce [

$/k

Wh

]

Normal TOU Non-CPP Day CPP Day

Off-Peak Part-Peak On-Peak Pa rt-Pe ak Off-

Modera tePrice

High Price

P art-P eak x 3

On-Peak x 5

8

Global Temperature Adjustment• Demonstrated large sheds in existing DDC EMCS

with minimal to no occupant issues • Comply with comfort standards• Oakland Federal Building, Sept. 8 2004

• Average of ~800 kW, 0.8 W/ft2 > 20% shed for 3 hrs. with two-step set point increase 72 F to 78 F

0

500

1000

1500

2000

2500

3000

3500

4000

45000:

001:

002:

003:

004:

005:

006:

007:

008:

009:

0010

:00

11:0

012

:00

13:0

014

:00

15:0

016

:00

17:0

018

:00

19:0

020

:00

21:0

022

:00

23:0

0

Who

le B

uild

ing

Pow

er [k

W]

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Pow

er In

tens

ity [W

/ft2]

Actual Baseline

ASHRAE 55-2004

Time Period2 ºF -15 min.3 ºF - 30 min.

4 ºF - 1 hr5 ºF - 2 hr6 ºF - 4 hr

9

• Alameda Count

2530 Arnold: Sep-22

0

50

100

150

200

250

300

350

400

450

500

0:0

0

1:0

0

2:0

0

3:0

0

4:0

0

5:0

0

6:0

0

7:0

0

8:0

0

9:0

0

10:0

0

11

:00

12:0

0

13

:00

14:0

0

15

:00

16

:00

17

:00

18

:00

19:

00

20

:00

21:

00

22

:00

23:

00

Wh

ole

Bu

ldin

g P

ow

er [

kW]

Actual LBNL Baseline CPP Baseline

IKEA: Oct-13

0

500

1000

1500

2000

2500

0:0

0

1:0

0

2:0

0

3:0

0

4:0

0

5:0

0

6:0

0

7:0

0

8:0

0

9:0

0

10:0

0

11:

00

12

:00

13

:00

14

:00

15

:00

16

:00

17

:00

18

:00

19:0

0

20:

00

21

:00

22

:00

23

:00

Wh

ole

Bu

ldin

g P

ow

er [

kW]

Actual LBNL Baseline CPP Baseline

Sample Automated CPP Response

10

Strategies at 32 DR Sites

CA-2003

CA-2004

CA-2005

NY

Glo

bal t

emp.

adj

ustm

ent

Fan-

coil

unit

off

SAT

res

et

Fan

VFD

lim

it

Duc

t sta

tic p

res.

res

et

Fan

quan

tity

redu

ctio

n

Ele

ctri

c hu

mid

ifie

r of

f

CH

W te

mp.

res

et

CH

W c

urre

nt li

mit

Chi

ller

dem

and

limit

Boi

ler

lock

out

Pre-

cool

ing

Ext

ende

d sh

ed p

erio

d

Slow

rec

over

y

Com

mon

are

a lig

ht d

im

Off

ice

area

ligh

t dim

Ele

vato

r cy

clin

g

Ant

i-sw

eat h

eate

r sh

ed

Foun

tain

pum

p of

f

Tra

nsfe

r pu

mp

off

Roc

k cr

ashe

rs o

ff

300 CapMall Office ● X X X X XACWD Office ● X X X X X X XAlbertsons Supermarket ● X XB of A Office ● ● ● X X X X XChabot Museum Museum ● X XCal EPA Office ● X X XCETC Office ● X XCisco Office/Data ● X X X X X2530 Arnold Office ● ● X X50 Douglas Office ● ● X XEchelon Office ● ● X X X X X XGSA 450 GG Office ● XGSA NARA Archives ● XGSA Oakland Office ● ● XGilead 300 Office/Lab ● XGilead 342 Office/Lab ● X XGilead 357 Office/Lab ● X XHome Depot Retail ● XIrvington High School ● X XIKEA Retail ● XKadent Industry ● XLafarge Industry ● XLBNL OSF Office/Data ● X XMonterey Office ● XNY Times Office ● X X X X X XOracle Office ● X XOSIsoft Office ● XRoche Office/Cafeteria ● ● XRockefeller Center Office ● X X XTarget Retail ● X XUCSB Library Library ● ● X X XUSPS Postal ● ● X X

Participation HVAC Light, Misc.

11

• Developed DR Control Categorization Framework

• Evaluated 32 Sites• Most successful strategy -

• Direct digital control global temperature adjustment

• In process for Title 24 2008• Closed loop

• Lighting Strategies - Zone Switching, Fixture Switching, Lamp Switching, Stepped Dimming, Continuous Dimming

• Need to link to Retro-Commissioning

Desire to tryDR

Global temp.adjustment

DDC zonecontrol?

Y N

Global temp.Adjustmentcapability?

Y N

Central plantcontrol

Airdistribution

control

Air distributionSystem DDC?

Y N

Can program GTA?

Y N Central plantDDC?

Y N

Do not try DRat this time

Desire to tryDR

Global temp.adjustment

DDC zonecontrol?

Y NDDC zonecontrol?

Y N

Global temp.Adjustmentcapability?

Y NGlobal temp.Adjustmentcapability?

Y N

Central plantcontrol

Airdistribution

control

Air distributionSystem DDC?

Y NAir distributionSystem DDC?

Y N

Can program GTA?

Y NCan program GTA?

Y N Central plantDDC?

Y NCentral plantDDC?

Y N

Do not try DRat this time

DR Control Strategy Categorization

Global Temperature Adjustment Thermal Mass Storage

Zone control

Reheat Loackout Duct Static Pressure Reset Fan VFD Limit Supply Air Temp Reset Fan Quantity Reduction

Air distribution

Cooling Valve Limit Chilled Water Temp Reset Chiller Demand Limit

Central plant

Chiller Quantity Reduction Recovery Slow Recovery Strategies

12

Links to Commissioning• Common question : If I can use a

strategy for demand response, why not use it all the time?

• Answer: Maybe you can!• Cx findings from Auto-DR buildings

• Air balancing• Duct static pressure reset• Zone temp reset• Night time fans left on

Continuum from energy efficiency, load management and demand response

13

Advanced Control Strategies:Shifting Demand with Pre-Cooling

Chabot Space and Science Museum

Single duct VAV/CAV systems

14

Chabot Museum Demand Shift Chabot: Whole Building Power, Sept 29

0 50

100 150 200 250 300 350 400

0:00 1:0

0 2:00 3:0

0 4:00 5:0

0 6:00 7:0

0 8:00 9:0

0 10:00 11:

00 12:00 13:

00 14:00 15:

00 16:00 17:

00 18:00 19:

00 20:00 21:

00 22:00 23:

00

Whole Building Power [kW]

Actual Baseline

75 kW shed

• Pre-cooled at 68 oF from midnight to 5am, • 70 oF from 5 am to 12 pm. • After 2 pm, temperature was gradually raised to 76 oF

15

Dynamic Zone Temperature Strategies (Chabot)

66

68

70

72

74

76

78

80

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Tem

p (

oF

)

current prclg+linear set up

prclg + agg linear set up ext prclg + exp set upNo prclg + linear set up

unoccupied hours occupied hours unoccupied

floating

floating

floatingfloating

floating

zonal reset

precooling

16

New York Times Building

• Technology designed for efficiency simulated to develop DR strategies

• Efficient features: Integrated movable, shading & dimming, Under floor air systems

• Commissioning in mockup• Demand Response Strategies

• Dimming lights beyond daylighting, reset zone temperatures (gradient), reduce perimeter fan speed, raise supply air temperature

Predicted Annual Savings from 400 kW Shed

Program Predicted Annual Savings*

Independent Capacity Program $17,632.00 Emergency DR Program $1,440.00 Distribution Load Relief Program $1,600.00

17

Future Control Systems: DR Modes in Control Architecture

• Orchestrate modes using schedules, signals, optimization algorithms:

• Occupied/Unoccupied• Maintenance/Cleaning• Warm up/Cool down• Night purge/Precooling• Low power DR mode

• Intelligence needed for decision making• Customized, simple and transparent interface• Financial feedback systems need to present

operational value• Similar capability for DOE’s Zero Energy Buildings• Embed DR Communications from EIS to EMCS

KWH

18

Collaborating with the CCC• Identify sites for PG&E Automated CPP

Project• Pursue commissioned sites• Pursue sites interested in Retro-Cx

• Provide review on DR Strategies Guide• Incorporate Manual or Automated DR in Cx

programs• Define DR modes and Cx modes in new

construction

19

Summary• DR capabilities improve with advanced controls that simultaneously support efficiency and

require good commissioning

• Field tests show DR potential 5-10% in most buildings with EMCS, yet limited knowledge of DR strategies: automation appears feasible with many existing systems.

• Need for Commissioning• Need DR control strategy commissioning tests• Need combined Retro-Commissioning DR Strategy Procedures

• DR is not driver, high performing buildings are:• Low energy costs, well-commissioned, low maintenance costs• Key is advanced controls, feedback systems, integrated performance

More information: drrc.lbl.govmapiette@lbl.gov