Demand Side Management in

Industry

Rangan Banerjee

Talk at Baroda in Birla Corporate Seminar August 31,2007

Demand Side Management
„ Indian utilities – energy shortage and peak
power shortage.
„ Supply Focussed- Build New Power Plants.
„ Capital & Fuel Scarcity, Environmental
Impacts, Gestation Period.
„ Need for Options – Demand Side
Management (DSM) & Load Management
 DSM Concept
„ Demand Side Management (DSM) - co-
operative action by the customer & the
utility(SEB) to modify the customer load
„ DSM benefits utility, consumer & society
„ Energy Conservation
„ Fuel Switching
„ Peak Clipping/Valley Filling/Load
Shifting
DSM terms
„ DSM Options- Options for customer load
modifications
„ Technological –energy efficient motors
„ Management- good housekeeping
„ Pricing – Time of Use Tariffs
„ DSM Programme = DSM option +programme
structure+ costs
„ DSM Plan – Sum of individual DSM
programmes
Electric Utility Role
„ Provision of Electricity to customers in
the service area- supply customers with
power they wish at whatever time
„ Planning under uncertainty of demand
„ Daily, Weekly & Seasonal Load Variation
„ Conventional power planning -Demand
– exogenous, uncontrollable – build
new supply to meet demand
Redefining Role
„ Capacity additions costly – Shortages of
peak power and energy
„ Low capacity utilisation of power plants
– increased costs – high tariffs
„ Gestation Period
„ Provision of energy services(using
electricity) to customers in the service
area (lighting, cooling, motive power…)
Utility Load Shape Objectives
„ Peak Clipping
„ Valley Filling
„ Load Shifting
„ Strategic Conservation
„ Strategic Load Growth
„ Flexible Load Shape
Peak Clipping

Peak Shortage
Energy Shortage
Valley Filling

Energy Surplus
during Off Peak
Load Shifting
Peak Shortage
Energy Surplus
Strategic Conservation

Energy Shortage
Strategic Load Growth

Energy Surplus
Flexible Load Shape
 DSM Programmes
„ Efficient Pumping Systems –Agricultural/
Municipal /Industry
„ Efficient Motor-Drive Systems - Industrial
„ Efficient Lighting - Commercial/Residential
„ Process Improvements- Industrial
„ Solar Water Heaters –
Residential/Commercial
„ Efficient AC – Commercial/ Residential
„ Cogeneration/Captive Power-
Industry/commercial
Inputs Required
„ Technology Characteristics
„ Base Case and DSM option – energy
consumption, peak coincidence
„ Capital, installation, O& M costs
„ Useful life
„ Present market share
„ Market Profile – market size – TM/RM,
market growth, diffusion rates
 Market Definitions

NEW
RM
Equipm Existing to be TM
Stock replaced

Existing Surviving

t=0 t=N
Technology Diffusion
„ Even if new technology better - will not
reach 100% acceptance (postponement
of acceptance, supply bottlenecks,
information gaps…)
„ Fisher-Pry model for substitution
„ df/dt = bf (1-f)
where f is the fractional market
share of the new improved technology
Fisher-Pry Model
„ ln (f/(1-f)) = a + bt
where a, b are constants
„ Blackman’s model - Final market share not
100% but F
„ ln (f/(F-f)) = a + bt
„ Determine a, b by method of least square
(regression) with initial substitution data
„ a,b by analogy
Fisher-Pry Curve
1

0.9

0.8

0.7

0.6
Fraction (f)

0.5 Series1

0.4

0.3

0.2

0.1

0
0 10 20 30 40 50 60
Time (Years)
Source: UNEP IRP Manual
(1997)
Cost Effectiveness of DSM
„ Perspectives
„ Utility
„ Ratepayer
„ Participant
„ Total resource/ Society
„ Net Present Value of Benefits – Costs, Cost of
Saved Energy (Rs/kWh), Cost of Saved
Demand (Rs/kW peak)
„ Different Discount rates
Programme Costs
„ Direct Costs- costs directly associated
with DSM option- include subsidy
provided by utility
„ Indirect Cost- Fixed + variable cost of
programme
„ Free Riders -Customers who would
have anyway adopted- yet utility pays
incentive (10-50% in US utilities)
Programme Costs
„ PC = I + A T + α C N
„ I - Initial Programme Set-up Cost
„ A- Annual recurring programme cost
„ α - proportion of cost sharing
„ N- number of adoptions
„ Energy Efficient motors I =Rs 20 lakhs,
A=Rs 3 lakhs, α =50% of
incremental cost
DSM Evaluation
„ Pre & Post programme metering
„ Analysis of Customer Billing data
„ Engineering analysis- physical based
models
„ Questionnaires & Survey
„ Statistical Modelling- demographics,
economics
End Use (HT Industry)
„ Motor/Pump/Fan 52 %
„ Air Compressor 9%
„ Air Conditioning/Refrig. 5%
„ Melting 16 %
„ Electrical Heating 11 %
„ Lighting 4%
„ Others 4%
 DSM Options

„ Energy Efficient Motors (EEM)

„ Variable Speed Drives (VSD)
„ Waste Heat Vapour Absorp. Refr. (VARS)
„ Industrial Cogeneration (COGEN)
„ High Efficiency Fans & Pumps (PUMPF)
„ Power factor correction (pf)
„ Electric Arc Furnace retrofit (EAF)
„ Efficient Lighting (ELB)/(HPSV)/(CFL)
„ Good Housekeeping(GHK)
„ Time of Use Tariffs (TOU)
 Technology Characteristics of
Motors

Range Typical STD Motor EEM Cost of Cost of

Rating hp Efficiency Efficiency Standard EEM
(%) (%) Motor (Rs)
(Rs)
1-5 3.0 79.7 86.8 7,500 9,750

5-10 7.5 84.4 88.6 13,300 17,290

10-15 12.5 87.3 91.0 24,100 31,330

15-20 17.5 88.4 92.0 28,500 37,050

20-50 35.0 90.6 92.0 56,200 73,060

>50 100.0 93.0 94.5 187,100 243,230

 No of Motors in each Category for HT
Industries in Maharashtra

Range 1-5 hp 5-10 hp 10-15 hp 15-20 hp 20-50 hp >50 hp Total

Percent of 15.3 10.8 11.1 9.9 13.8 39.1 100.0

motor
connected
load

Connected 287.2 202.7 208.3 185.8 259.0 733.9 1877.

MW 1989-90 0

‘000s units 128.3 36.2 22.3 14.2 9.9 9.8 220.7

1989-90
‘000s units 157.2 44.3 29.3 17.4 12.1 12.0 272.3
1992-93
 DSM HT Industrial Plan Results Maharashtra

DSM DEMAND ENERGY PROG. COST U T IL IT Y CSE

O P T IO N (M W ) (G W h ) (M IL L IO N R s ) R s /k W p /k W h
PF 4 0 .2 - 41 1000 -
TOD 1 1 0 .4 - 190 1700 -
EAF 1 7 .8 9 4 .9 36 2000 20
CFL 1 .1 4 .7 3 2900 61
GHK 5 5 .2 2 2 8 .8 208 3800 86
HPSV 1 .4 7 .2 9 6500 -1 0
PUMPF 1 6 .1 8 0 .5 140 8700 77
EEM 9 .3 4 6 .4 83 9000 63
VSD 3 7 .4 3 3 3 .1 381 10200 105
VARS 1 1 .2 7 9 .1 119 10600 64
ELB 2 .2 9 .8 28 12400 100
TOTAL 3 0 2 .3 8 8 5 .3 1238 4100 82
COGEN 2 4 2 .3 1 3 5 8 .0 1162 4800 76
TOTAL 5 4 4 .6 2 2 4 3 .3 2400 4400 78
Least Cost Curve for DSM
14000

12000

10000
Cost of Demand Saved (Rs/kW)

8000

6000

4000

2000
Pf

0
0 50 100 150 200 250 300

D e m a n d S a v in g (M W )
Implementation of Industrial DSM?
„ Technology available, yet low
implementation
„ SEB interested only in peak reduction,
not energy conservation
„ Power factor Correction - for industrial
penalty/incentives
„ Analysts conclude SEBs supply focussed
Utility Viewpoint
„ ROR clause on net fixed assets
„ Strategy - Increase asset base -build
new power plants
„ DSM Investment - on Clients site -no
asset generation for SEB
„ No idea of transaction costs
„ Revenue Considerations
 Consumption Profile for MSEB
(1992-93)
Consumer Class No. of Energy % of Tariff/
Consumers Consumption Sales Cost of
kWh/Yr/Consumer Supply
Domestic 5.5 million 600 13.3 0.5

Commercial 1.4 million 2,100 6.5 1.4

Agricultural 1.8 million 6,200 21.1 0.1

Industrial (low tension) 0.3 million 11,100 7.1 0.7

Industrial (high tension 7000 2 million 33.2 1.8

 Factors Affecting DSM
Resid Agric Comm Ind
Revenue + ++ - --
Impact
Savings + ++ + ++
Potential
Transaction -- -- -+ ++
Cost
Customer - -- ++ ++
Viability
 Viability of Compact Fluorescent Lamps
10

8 1000 hrs
Cost of Saved Energy (Rs/kWh)

4
4000 hrs

0
0 .0 0 .2 0 .4 0 .6 0 .8 1 .0

D is c o u n t R a te ( d )
Survey results (DSM options)
Ranking of DSM Options(Survey)
 From Jointly Owned

Energy balance
WIND
State Projects and
500 MU Captive
Central Sector
GAS ( 1 %) Generation
19500 MU
5450 MU 4500 MU
HYDRO ( 8 %)
STEAM 5600 MU
55500 MU (8 %)
(83 %)

TOTAL 24000
MU

STATE GENERATION TOTAL 67050 MU

GRAND TOTAL 91050 MU

AUX.CON.
4780 MU
NET POWER
SELL TO OTHER
59620 MU STATES
(65 %) 150 MU

PUBLIC
T & D LOSS GROSS ENERGY CONSUMPTION 59620 MU LIGHTING
26500 MU 640 MU
32 % WATER WORKS &
PUMP 1540 MU (3 %)

TRACTION
1850 MU
AGRICULTURAL ( 3 %)
INDU. 10730 MU
POWER RESIDENTIAL (18 %)
COMMERCIAL
(LV) 12660 MU
INDUSTRIAL 5350 MU
4820 MU (21 %)
POWER (HV) (9 %)
22030 MU ( 8 %)
(37 %)
 Generalised lighting end use model

Framework of model proposed for construction of end use load profiles

Economic analysis from different perspectives
ƒ Economic analysis from perspective of Utilities
3 3

Average Energy Purchase Cost in Rs/kWh

2.5 2.5
Cost of Saved Energy in Rs/kWh

Energy Purchased by MSEDCL

Highest Power generation cost of MSPGCL = 1.81 Rs/kWh

2 2

1.5 1.5

Energy Saved by Lighting Energy Efficiency

1 1

0.5 0.5

0 0
0 500 1000 1500 2000 2500 3000 3500 4000
Energy Saved/ Purchased in GWh
Load Management Steps
„ Analysis of Load variations
„ Identification of Controllable
Loads
„ Selection of Control Option

„ Implementation strategy
 Common LM Options (SEB)
„ Staggering of working hours of large
consumers
„ Staggering of holidays of large consumers
„ Specified energy and power quotas for major
consumers
„ Rostering of agricultural loads
„ Curtailment of demand - service interruptions
(load shedding)
Quantification and trend of energy demand

Variation of monthly average energy demand for MSEDCL

Quantification and trend of peak demand
ƒ Construction of load duration curve
ƒ Provides an idea about duration of the time when system load is
> specified load

A Sample Load Duration Curve

Quantification and trend of peak demand

Quantification of peak shortage of MSEDCL for 2006-7

Load Management Options
„ Direct Load Control (DLC) – Utility has
control of directly switching off
customer loads
„ Interruptible Load Control (ILC)- Utility
provides advance notice to customers
to switch off loads
„ Time of Use (TOU) Tariffs – price signal
provided – customer decides response
US DLC Programmes
„ 5352 MW of peak reduction in US in
1995 by DLC (11,237 MW of potential
peak reduction)
„ Electric water heaters, Air-conditioners
and space heating systems
„ Radio Control (80% of DLC)
„ Power Line Carriers
„ Ripple Control
DLC Control Strategies
„ Cycling - Groups of loads switched off
for short time periods to reduce system
diversified demand
„ Payback control- Groups of devices
switched off for periods upto 6 hours
„ Cycling preferable as payback control
imposes larger than normal load when
switched on
LM Options
„ Cool Storage – Chilled water , Ice
storage- operate compressor during off-
peak
„ Water pumping systems
„ Cogeneration – Operating strategy
„ Evaluate Process Storage possibilities
„ Power pooling
DLC- India
„ Central Air Conditioning- cities like
Delhi, Mumbai … VHF controlled or
timer controlled - cycling
„ Agricultural pumping
„ Reconfiguration of Distribution system
„ Individual controllers on pumps
„ Timer Control
„ Municipal Water Pumping
 Interruptible Load Control
„ Notification Period
„ Minimum Interruptible Load
„ Incentive (Rs/kVA/month of IL)
„ Limit on no of interruptions
„ Group Load Control
„ US 1995 - 8401 MW actual reduction
(21820 MW PPR)
Indian ILC
„ Voluntary
„ Large Industrial & Commercial
Consumers with minimum interruptible
demand of 500 kW
„ Choice of notification period (24 hours
or 2 hours)
„ Pilot programmes
„ Monitoring of load curves
 A n alysis of S ystem load cu rve

Id en tification of p eak, p artial p eak Flow Chart for

an d O ff p eak p eriod s fo r th e system

TOU Tariff
In stall tim e of u se (T O U ) m ete rs fo r la rge
in d u strial & com m e rcia l con su m e rs

A n alyse load p ro files

E stim ate th e p oten tial of load sh iftin g fo r

d iffe ren t p ricin g con d ition s

S et u p exp e rim en ta l T O U tariff

E valu ate Im p acts of T a riffs

Is a revision of
tariff req u ired ?

Y es

T O U tariff

No
TOU Tariff
„ UK study showed that TOU only feasible
for consumption > 3 MWh/year per
consumer
„ Time differentiated Energy charges
„ Time differentiated Capacity charges
„ Rating periods - 3 to 4 for simplicity
„ Spot Pricing ?
 Time of Use Tariff (MSEB-HT Ind., Jan 2002)

450

400

Peak
Peak
350

300

Partial
Paise/kWh

250

Peak
200
Off-peak
150

100

50

0
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
Hours
Importance of Energy Cost
„ Energy cost- Significant component of
manufacturing cost
„ Energy prices – Higher growth rates
than other inputs
„ Higher Input energy prices vis-à-vis rest
of the world
„ Competitive Disadvantage
Strategies for Electricity Bill
Reduction
„ Cross Subsidy – Tariff 1.8 times cost of
supply
„ Electricity Bill = Average Electricity Price ×
Electricity Consumed
1. Energy Efficiency Options to reduce electricity
consumption
2. Reduce the average electricity price – by opting for
cogeneration
3. Reduce average electricity price –by analysing tariff
structure and modifying usage pattern.
Energy Efficiency
• Determine end-use profile of plant load
• Motors (80% of load), lighting, heating
•Breakup of motor load for a Chemical plant

Conveyors Others
(5.5 %) (15.9 %)

Centrifuges
(1.3 %)
Pumps
Agitators
(51.3 %)
(12.7 %)

Fans
Compressors (11.9 %)
(1.3 %)
Energy Efficiency
„ Determine operating efficiency of motors
Portable power meter, tachometer
„ Motor audit – identify lightly loaded,
inefficient motors – replace with Energy
efficient motors.
„ Pumping – VSD, trimming impeller
„ Compressed Air – reduce pressure, air leakages
„ FRP fans – Cooling towers, energy efficient
fans
„ Waste heat - Vapour Absorption Refrigeration
„ Energy efficient process, SEC norms
 DEFINE AUDIT O BJECTIVES

Q UESTIO NNAIRE

REVIEW PAST RECO RDS

W ALK THRO UG H /
PLANT FAM ILIARISATIO N

DATA REQ UIREM ENTS

DATA
ANALYSIS

M EASUREM ENTS / TESTS

CO M PUTE M ASS / ENERG Y

BALANCES

INSTALL
M EASURES
ENUM ERATE ENERG Y
CO NSERVATIO N O PPO RTUNITIES

EVALUATE ECO s

PRIO RITISE RECO M M ENDATIO NS

 Comparison of initial cost and life cycle cost

Annual Cost of electricity

Initial cost
S. No. Equipment Rating Electricity ALCC (Rs) as %
(Rs)
Cost (Rs) of ALCC

1 Motor 20 hp 45,000 600,000 605,720 99.0

2 EE Motor 20 hp 60,000 502,600 512,700 98.0

3 Incandescent 100 W 10 1168 1198 97.5

Lamp

4 CFL 11 W 350 128 240 53.6

EE- Energy Efficient, CFL- Compact fluorescent lamp, ALCC- Annualised life cycle cost
Sample Industrial Load Profile (Mumbai)
ILM Research Objective
„ Determine optimal response of industry
for a specified time varying tariff –
develop a general model applicable for
different industries
„ Process Scheduling- Continuous/ Batch
„ Cool Storage
„ Cogeneration
Process Scheduling
„ Variable electricity cost normally not included
„ Flexibility in scheduling
„ Optimisation problem – Min Annual operating
costs
„ Constraints – Demand, Storage and
equipment
„ Models developed for continuous and batch
processes (Illustrated for flour mill and mini
steel plant)
„ Viable for Industry
Process Scheduling
„ Batch processes- batch time, quantity,
charging, discharging, power demand
variation (load cycles)
„ Raw material constraints, Allocation
constraints, Storage constraints,
Sequential Constraints, maintenance
downtime
Process Scheduling Summary
Example Structure Results Saving

Flour Mill Linear, IP Flat- 2 1%

Continuous 120 variables shift - 6.4%
46 constraints 25%store 75%peak
TOU-3 reduction
shift
Mini Steel Linear, IP Flat 8%
Plant 432 variables TOU 10%
Batch 630 Diff 50% peak
constraints loading reduction
Steel Plant Flow Diagram
St. steel Scrap mix or Bloom caster
Alloy steel scrap mix VD or VOD
station

Ladle Arc
40 T Melting Arc
furnace Open store
furnace

Convertor (only for

Alloy steel St Steel) Reheat furnace
scrap mix
Billet caster

Bloom mill

Open store
30 T MeltingArc furnace
Open store

Reheat furnace

Bar mill

Open store ooo

Wire products
ooo
for final finish
Reheat Wire mill
furnace

Rods, Bars for final

finish
 Steel Plant Optimal Response to
TOU tariff
60 Optimal with TOU tariff
Optimal with flat tariff
50

40
Load MW

30

20

10
0
2 4 6 8 10 12 14 16 18 20 22 24
Time hours
Cool Storage
„ Cool Storage – Chilled water operate compressor
during off-peak
„ Commercial case study (BSES MDC), Industrial
case study (German Remedies)
„ Part load characteristics compressor,pumps
„ Non- linear problem – 96 variables, Quasi Newton
Method
„ MD reduces from 208 kVA to 129 kVA, 10%
reduction in peak co-incident demand, 6% bill
saving
Schematic of AC Plant with Cool storage

Cooling Tower Condenser

pump

Condenser
Expansion
device

AHU
Space to be Compressor
conditioned
Chiller
pump
Chiller

Additional
Chiller pump
Under ground
Chilled water
Storage tank
 Cooling Load and Electric Demand Profiles
without storage

Co o lin g Lo a d
250 MD 207.8 kW E le c tric De ma n d

200
kW, TR

150
100 Max.170 TR

50
0
1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Time hours
 With Cool Storage
f lat tar if f

250 20 7. 8 kW TOU tar if f

200 Exis ting without cool

s tor age
Load kW

150 1 22. 5 kW

100 121 . 7 k W

50
0
1 2 3 4 5 6 7 8 9 101112131415161718192021222324
Time hours
Cogeneration
„ Process Steam, Electricity load vary with time
„ Optimal Strategy depends on grid
interconnection(parallel- only buying,
buying/selling) and electricity,fuel prices
„ For given equipment configuration, optimal
operating strategy can be determined
„ GT/ST/Diesel Engine – Part load characteristics –
Non Linear
„ Illustrative example for petrochemical plant-
shows variation in flat/TOU optimal.
 Stack
Fuel, HSD WHRB-1
5.9 T/h G
1
136 T/h Cogeneration Example
20 MW
Gas turbine -1 Supp. Firing
LSHS 5.6 T/h
Grid Process Load,
Fuel, HSD WHRB-2 136 T/h 7.52 MW 60 MW
5.9 T/h
G
1
Gas turbine -2 20 MW Supp. Firing BUS
LSHS 5.6 T/h
SHP Steam 100 bar,500o C

60.6 T/h
131.7 T/h 117.1 Process Load,150 T/h 76.2 T/h
T/h 12.5
Boiler
MW

ST
Fuel, LSHS Process Load,125 T/h G4
PRDS-1
9.64 T/h PRDS-2
HP Steam 41b,400 oC 20 T/h

MP Steam 20b, 300 oC

Process Load 40 T/h 49.5 T/h 40 T/h 16.2 T/h

PRDS-3
LP Steam 5. 5 b, 180 oC
53.4 T/h

Process Load Deaerator

40 T/h Feed water Condenser
426.5 T/h Make up water,357 T/h
 Import Power from Grid with
Cogeneration for a Petrochemical Plant

flat tariff TOU tariff

Import power MW

25 21.6
20 17.6 peak
period
15 demand
10
11 MW
5
0
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

Time hours
 Export power to the grid with
Cogeneration for a Petrochemical Plant
Peak
Export Power MW

40 period
flat tariff TOU tariff
30 demand

20 9.7 MW
10

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

Time hours
- Integrated approach
Optimal Operating
Process demand profile, Captive/Cogeneration operating cost
Cooling electric load power model strategy of structure
profile, Steam load profile captive/
cogeneration
Grid tariff, fuel costs, plant
Grid conditions

Process load Plant

model /measured
Air conditioning
input data
(cooling) load model

Optimal process load

schedule Optimal cool storage

Modified process
demand profile
Modified cooling electric
load profile
Modify steam load
profile for process
related loads
Establish norms
„ Metering/Sub-metering
„ Energy consumption vs load
„ Specific energy consumption (SEC)
lower at full capacity
„ Compare with other plants/ Best
practice plants
„ Energy audits every year
 SEC for Furnaces

Specific energy consumption 12000

10000
(SEC) (kJ/kg)

8000

6000

4000

2000

0
0 20 40 60 80 100
Cumulative % furnace

Theoretical SEC SEC for Indian glass furnaces

SEC for 123 glass furnaces*

*Rund G. C., ‘Energy Efficiency Benchmarking of Glass Furnace’,

62ed Conference on Glass Problems, University of Illinois , Urbana Champaign , 2001
Extrapolation to actual furnaces
Concluding Remarks
„ DSM for electricity bill reduction
„ Carbon Dioxide reduction / CDM
„ Cogeneration/ Trigeneration
„ Process Scheduling
„ Benchmarking/ Target setting
„ Competitive Advantage
 References
„ J.K. Parikh , B.S.Reddy and R.Banerjee, Planning for Demand Side Management
in the Electricity sector, Tata McGraw Hill , New Delhi,1994.
„ S Ashok and Rangan Banerjee, Optimal Operation of Industrial Cogeneration for
Load Management, IEEE Transactions on Power Systems, Vol 18, No. 2,
May 2003.
„ S. Ashok and Rangan Banerjee, Optimal cool storage capacity for load
management, Energy 28 (2003) 115-126.
„ R.Banerjee, "Load management in the Indian power sector using US experience",
Energy, Vol 23, 1998 , pp 961-973
„ J. K. Parikh, B. S. Reddy, R. Banerjee & S. Koundinya," DSM survey in India:
awareness, barriers and implementability", Energy, Vol 21, No 10, pp 955 -
966, 1996.
„ P.Du Pont et al, Lessons learnt from implementation of DSM in Thailand, Energy
Policy,1998
„ V.Sardeshpande, U.N. Gaitonde and R BanerjeeModel based energy
benchmarking for glass furnace Energy Conversion and Management, In
Press, Available online 19 June 2007,