modelling of energy systems- renewables and efficiency rangan banerjee department of energy science...

Modelling of Energy Systems-Renewables and Efficiency

Rangan BanerjeeDepartment of Energy Science and

Engineering IIT Bombay

Seminar at Strathclyde University, June 2, 2009

Indian Examples

Analysis of wind integration into power system

PV- Battery storage – Design space Solar Water Heater Diffusion Optimal response to time of use tariff - process scheduling - cool storage -cogeneration Benchmarking of glass furnace Decision Support System for energy planning

Issues in grid integration

Conventional power planning- hydro-thermal scheduling

How do we deal with renewables? Capacity credit New methodology based on Load

Duration Curve Illustrated for Wind in Tamil Nadu

Tamil Nadu location

Source Installed capacity (MW)

Annual Energy

generated(MU)

Annual average capacity

factor (%)

Coal 2970 21230 81.6

Gas 424 1945 52.4

Hydro 2187 6290 32.8

Firm central share 2825 17785 71.9

Wind (state + private) 3856 5270 18.6

Other renewables (solar PV, biomass and Bagasse based

cogeneration)556 1220 25.1

Independent power projects (coal, lignite, diesel or gas

based)1180 6360 61.5

Assistance from other regional grids

519 2280 50.1

Total 14517 63370 49.8

Tamil Nadu – Grid Details

TN – Installed wind power and wind energy generated

0

500

1000

1500

2000

2500

0 4 8 12 16 20 24

Hours

Po

wer

gen

erat

ed in

MW January

June

September

Mean value

0

200

400

600

800

1000

1200

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Months

Win

d e

ne

rgy

ge

ne

rate

d (

MU

)

Hourly variation of wind power

Monthly variation of wind energy generated

Capacity Credit Methodology

Mean Value

0

5

10

15

20

25

30

35

40

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Months

Cap

acit

y cr

edit

(%

)Variation of CC

0

5

10

15

20

25

30

35

40

0 5 10 15 20 25 30 35 40

Monthly capacity factor (%)

Mo

nth

ly c

apac

ity

cred

it (

%)

Percentage penetration of wind power (%)

Wind power installed capacity (MW)

Capacity credit (MW)

0 0 0

5.5 500 130

11.1 1000 240

16.7 1500 350

22.2 2000 460

33.3 3000 675

44.4 4000 895

61.1 5500 1220

77.8 7000 1550

88.9 8000 1750

100 9000 1965

Methodology – Wind (Micro level)

Select major sitesHourly wind speed data at sensor height

Wind turbine characteristics

Extrapolated hourly wind power generated in the

state

Installed capacity at each site

Hourly wind power at sensor height

Hourly wind power at hub height

Hub height, power law index at each site

Continue to LDC Methodology

Input n and n discrete wind capacities

Select major sites

Extrapolated hourly wind power generation

Effective load curve

Divide load curve into 100 MW bins

Record number of hours in each bin

Calculate effective base and peak load savings from different LDCs obtained

Evaluate for n discrete wind capacities

Sum up to obtain annual load duration curve

Frequency distribution of load over the year

Hourly wind

speed data

Wind turbine

characteristicsInstalled

capacity of wind power

Hourly load

curve

Impacts on LDC

Wind power installed capacity (MW)

Base load capacity saved (MW)

Peak load capacity saved (MW)

0 0 0

500 60 70

1000 100 355

1500 150 1105

2000 240 1265

3000 470 1475

4000 770 1625

5500 1150 1775

7000 1460 1975

8000 1630 2085

9000 1855 2125

LDC Methodology for future scenarios

Present LDC

GDP growth rate

Elasticity

Projected LDC

Projected year

Projected LDC with renewables

Projected renewables installed capacity

Evaluate of discrete renewable energy capacities

Compute base and peak capacity saved

TN – Wind energy scenarios for 2021

Simulation for UK

Site: Valley, Hollyhead, Anglessy (Wales)

Mean monthly wind speed

0123456789

1 2 3 4 5 6 7 8 9 10 11 12month

win

d sp

eed

(m/s

)

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20 25

Hrs.

no

rma

lise

d p

ow

er

ge

ne

rate

d

UK load

wind power

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20 25

Hrs.

No

rmal

ised

w.r

.t.

pea

k

Load

wind power

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20 25

Hrs.

No

rmal

ised

gen

erat

ion

w.r

.t.

pea

k Load

Wind power

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20 25

Series1

Series2

Wind power - load curve correlation in UK

July – 0.88 Jan – (-)0.51

Mar – 0.48 Oct – 0.39

Average correlation factor over the year = 0.38

No wind power

3200 MW wind (present installed capacity)

6000 MW wind (2011 estimate)

Results for UK

Installed solar power capacity

Hourly total renewable energy generated

Installed wind power capacity

Installed biomass power capacity

Hourly solar energy generated

Hourly wind energy generated

Hourly biomass energy generated

Effective load curve

Divide load curve into equal-sized bins

Record no. of hrs. in each bin

Sum up to obtain annual LDC

Frequency distribution of load over the year

Evaluate for n discrete renewable energy capacities

Hourly load curve

Compute base and peak load savings from diff. LDCs

Input n and n discrete capacities of W, S and B

Wind resource model

WECS performance model Solar PV performance model Biomass power generation

Solar resource model

Utility generation model Load model

Economic scenarioCapacity expansion model

Output: Capacity Savings

Micro level model

Macro level model

TN Solar Methodology (Micro-level)

Select major sites

Hourly solar power generation at each site

Correction for panel inclination

Correction for temperature effect

Extrapolate based on installed capacity at

each site to get hourly solar power generation

for the state

Continue to LDC Methodology

Hourly average insolation

Solar module characteristics (efficiency vs insolation)

Power coefficient (W/deg. C)

Cos ()

100

125

150

175

200

225

250

jan feb mar apr may jun jul aug sep oct nov dec

months

Mo

nth

ly g

en

era

tio

n f

or

10

00

MW

so

lar

PV

(M

U)

Solar power – monthly variation

0

20

40

60

80

100

120

140

160

180

jan feb mar apr may jun jul aug sep oct nov dec

months

en

erg

y g

en

era

ted

in

MU

Biomass power – monthly variation

Installed capacity – 450 MW (340 MW from bagasse cogen)

Wind and solar power required for 1000 MW avg. peak saving

y = -0.8666x + 2798.1

R2 = 0.9834

0

500

1000

1500

2000

2500

3000

3500

0 500 1000 1500 2000 2500 3000 3500 4000

Installed wind power (MW)

Ins

talle

d s

ola

r p

ow

er

(MW

)

Wind and solar power installations to replace 1000 MW base power

Slope = 0.8666 which is almost equal to ratio of capacity factors (0.18/0.21 = 0.857)

Base capacity savings with wind + solar

Peak capacity savings with wind + solar

Hybrid scenarios – Impacts on LDC

A

B

C

Impacts on LDC – capacity savings

Scenario Wind (MW)

Solar (MW)

Biomass (MW)

Base capacity saved (MW)

Base capacity saved (% of installed RE capacity)

Peak capacity saved (MW)

Peak capacity saved(% of installed RE capacity)

A 4000 1000 1000 1568 30.31 1252 27.22

B 5000 1000 1500 2137 28.66 1423 26.79

C 3000 3000 3000 2984 33.15 2276 25.29

Design approach

Case study

of sample systems

Analysis

Decision

making

Integrated designmethod

Sample design

INPUTOUTPU

T

Develop generic

guidelines for

design

Compare with

Existing design

methods

Integrated design of Isolated power system

Load estimation

Sizing

Distribution networkusing ViPOR

Load flow analysis

Is the current location of

source gives minimum loss

Relocate the source

No

Yes

End

Name of the plant

ConnectedLoad (kW)

PlantCapacity

Distribution loss (%)

Plant capacity factor (%)

Energy costRs / kWh

Existing Designed Existing Designed Existing Designed Existing Designed

Solar PV, Rajmachi

1.4 5 kWp 4 kWp 4.6 0.5 8.3 11.5 32 25

Biomass gasifier, Dissoli

6.9 10 kW 10 kW 12.3 2.0 8.8 12 29-37 21-25

Biomass gasifier, Lonarwadi

10.7 20 kW 10 kW 14.6 2.7 5.6 14 43-54 16-25

Integrated design-Summary

Photovoltaic array

AC busDC bus

Charge controller

Battery bank

Inverter Load

Sizing of Photovoltaic-Battery Systems

Objective:

To arrive at the set of all feasible configurations (Array rating and Battery capacity) to meet a given demand following a time-series simulation of the system

Schematic of the System

Mathematical model

Energy balance

Repeatability of battery energy

Non negativity of battery energy

For a small time step, battery energy:

Hourly energy balance

Battery storage requirement

Power from the photovoltaic array

Photovoltaic-Battery System Sizing (Deterministic Approach)

Inputs: Hourly solar insolation data, Hourly load data,

Photovoltaic system efficiency, Power conversion efficiency

Estimation of the solar insolation incident on the array

System simulation to obtain the minimum array size and the corresponding battery capacity

Calculation of the minimum storage capacity for different array sizes greater than the minimum

Plot of sizing curve and the identification of the

design space

Array rating (W)

Bat

tery

cap

acity

(Wh)

Infeasible design region

Minimum arraysize

Feasible design region(design space)

sizing curve

Battery capacity corresponding to the minimum array size

Graphical representation Sizing curve for given solar insolation profile, load curve and system characteristics

0

5

10

15

20

25

30

0 2 4 6 8 10 12 14 16 18 20 22 24

Time (hour of the day)

Lo

ad

(kW

)

0

100

200

300

400

500

600

700

0 2 4 6 8 10 12 14 16 18 20 22 24

Time (hour of the day)

Inso

latio

n (

W/m

2)

Photovoltaic-Battery System Sizing (Example)

Solar insolation profile

Averaged values for the month of December

Demand profile

For an average day

For a remote location in Sagar island, West Bengal

238

240

242

244

246

248

250

30 35 40 45 50 55

Array power(kWp)

Ba

ttery

ca

pa

city

(K

Wh

)

Minimum array rating

Sizing curve

Feasible region

Infeasible region

Sizing curve and Design-space for the example

Mathematical Formulations Chance constraint:

Incorporating energy conservation equation of the storage:

Deterministic equivalent:

Generation of design space incorporating other constraints

)}()({ tDtDyProbabilit actual

)()()(

)(

)(

)(tPtD

ttf

tQ

ttf

ttQyProbabilit actual

BB

))(()(

)(

)(

)()()( ztD

ttf

tQ

ttf

ttQtPactualtP

BB

PV-Battery System

0.6

0.8

1

1.2

1.4

0 1 2 3 4 5 6 7

Array rating (kWp)

Bat

tery

cap

acity

(kW

h)

α = 0.5

α = 0.7

α = 0.8

α = 0.95

Solar Water Heaters

Estimate potential for solar water heaters in a given area Develop generic framework ‘Diffusion of Renewable Energy Technologies’

Factors Affecting Diffusion Of SWHS

Location- Insolation Water Usage Pattern Cost of electricity Capital Cost Reliability Potential savings Subsidies/ Financial Incentives

Micro Level Decision Model (Parametric Analysis)

TRNSYS

INPUT DATAWater usage pattern

Location (Monthly average hourly temperature and radiation data)

Characteristics of SWHS Auxiliary heating requirement (Monthly average hourly data)

Economic Analysis MS EXCEL

Savings in Electricity Cost Payback Period Analysis Cost of electricity saved

Selection and sizing of SWHS

TRNSYS (Transient System Simulation Program developed at SEL, University of Wisconsin)

Information Flow Diagram of Micro –simulation for SWHS

Weather data

COLLECTOR

SOLAR RADIATION PROCESSOR

STORAGE TANK

LOAD(Hourly hot water

usage pattern)

AUXILIARY HEATER

Hourly Global Solar Radiation& Diffuse Solar Radiation

Hourly Solar Radiation on Collector SurfaceHourly ambient

Temperature

Auxiliary heating

requirement

Target areaWeather data, area details

Identification and Classification of different end uses by sector (i)

Residential (1)Hospital (2) Nursing

Homes (3)Hotels (4)

Others (5)

POTENTIAL OF SWHS IN TARGET AREATechnical Potential (m2 of collector area)

Economic Potential (m2 of collector area) Market Potential (m2 of collector area) Energy Savings Potential (kWh/year) Load Shaving Potential (kWh/ hour for a monthly average day)

* Factors affecting the adoption/sizing of solar water heating systems

Sub-class (i, j)

Classification based on factors* (j)

Single end use point

Potential

Base load for heating

Electricity/ fuel savings

Economic viability

Price of electricity

Investment for SWHS

Technical Potential

SWHS capacity

Constraint: roof area availability

Capacity of SWHS

(Collector area)

TargetAuxiliary heating

Single end use point

Micro simulation using TRNSYS

Hot water usage pattern

Weather data

SIMULATION

Auxiliary heating requirement

No. of end use points

Technical Potential

Economic Potential

Economic Constraint

Market Potential

Constraint: market acceptance

Potential for end use sector (i = 1) Potential for i = 2

Potential for i = 3

Potential for i = 4

Potential for i = 5

Model for Potential Estimation of Target Area

Potential Of SWHSTechnical potential Pij for sub-class j in sector i is

where j denotes sub-class of end use points in sector i.

Psj is the simulation output for a single end use point

fj denotes fraction of the end uses m is the total number of sub-classes.

faj is fraction of roof area availability

Ni is the number of end uses points in sector i

Technical Potential for sector i is

where i denotes sector

Technical Potential of SWHS P(T) in the target area is

sjPiNajfjfijP

m

1jijPiP

iP)T(P

Economic Potential

Economic potential of SWHS P(E): subset of technical potential

ve = 0, if payback period > maximum acceptable limit

ve = 1, if payback period < maximum acceptable limit

ijij PEP ev

Payback Acceptance Schedule

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 2 4 6 8 10 12

Payback period (years)

Fra

ctio

n M

eeti

ng E

cono

mic

C

rite

ria

MARKET POTENTIAL

fp,j is fraction of potential adopters meeting economic criteria.

ijPpjfijMP

Input Data For Potential Estimation Of SWHS in Pune

Target Area Pune

Area 138 sq.km

Total Number of households 5.17 lakhs

Number of households with more than three rooms 1.41 lakhs

Average number of persons in each household 5

Number of hospitals 394

Capacity range of hospitals 1-570 beds

Number of nursing homes 118

Capacity range of nursing homes 1-50 beds

Number of hotels 298

Capacity range of hotels 10-414 inmates

Number of households residing in single ownership houses 35250

6 floors 1400

10 floors 880 Number of buildings (4 flats in each floor)

11 floors 840

Residential 2.80 Cost of electricity (Rs./kWh) Commercial 4.00

Hot Water Usage Patterns (Pune)

(a) Residential (1) [Gadgil, 1987]

0

10

20

30

40

50

60

70

0 2 4 6 8 10 12 14 16 18 20 22 24Hour of day

litr

es/

hTemperature = 40 o C

(b) Residential (2) [Narkhede, 2001]

020406080

100120140160180200

0 2 4 6 8 10 12 14 16 18 20 22 24Hour of day

litr

es/

h

Temperature = 40 o C

(c) Hospital (1 bed)

0

5

10

15

20

25

0 2 4 6 8 10 12 14 16 18 20 22 24Hour of day

litr

es/

h

Temperature = 50 o C

(d) Nursing Home (1 bed)

0

2

4

6

8

10

12

0 2 4 6 8 10 12 14 16 18 20 22 24Hour of day

litr

es/

h

Temperature = 50 o C

(e) Hotel - 1 guest

0

5

10

15

20

25

30

0 2 4 6 8 10 12 14 16 18 20 22 24Time of day (Hour)

itres/

h

Temperature = 60 o C

Monthly Average Ambient Conditions in Pune

0

5

10

15

20

25

30

35

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Month

Mon

thly

Ave

rage

Sol

ar

Rad

iati

on (

kW

h/m

2 /day

)

0

5

10

15

20

25

30

35

Mon

thly

Ave

rage

Am

bie

nt

Tem

per

atu

re (

o C)

Incident Solar Radiation

Ambient Temperature

Mani, A. (1980) ‘Handbook of solar radiation data for India’.

Sample simulation output and potential estimation for hospital with 5 beds

(a) Energy flow/ Solar Radiation for a typical day

0

2000

4000

6000

8000

10000

12000

0 2 4 6 8 10 12 14 16 18 20 22 24

Hour of day

Solar RadiationSolar EnergyAux.heating

(b) Temperature profiles for a typical day

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12 14 16 18 20 22 24

Hour of the day

Amb. Temp.Temp.at collector outletTemp. at tank outletTemp. at load

(c) Monthly variation in electricity savings

0

50

100

150

200

250

300

350

400

450

Month of year

Annual Electricity Savings = 4290 kWh

Potential Estimation of Sectors (Pune)

Technical Potential Market Potential

Sector Collector area (m2)

Annual Electricity

savings (kWh)

Collector area (m2)

Annual Electricity

savings (kWh)

Single houses 106000 37200000 2100 740000 Residential

Multi-storeyed 227400 165000000 41000 29700000

Hospitals 5500 5900000 1700 1600000

Nursing homes 600 500000 300 280000

Hotels 13800 15900000 9300 10740000

TOTAL 353300 224500000 54400 43100000

Load Curve Representing Energy Requirement for Water Heating

0

100

200

300

400

500

600

700

800

900

1000

0 2 4 6 8 10 12 14 16 18 20 22 24Hour of day

Ene

rgy

Con

sum

ptio

n (M

W)

Typical day of January

Typical day of May

Total Consumption =760 MWh/day

Total Consumption = 390 MWh/day

53%

Electricity Consumption for water heating of Pune

Total Consumption =14300 MWh/day

Total Consumption = 13900 MWh/day

Total Electricity Consumption of Pune

Achievable Potential Of SWHS For Different Payback Periods (Pune)

0

50000

100000

150000

200000

250000

300000

350000

400000

0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00

Payback Period (years)

Tec

hnic

al P

oten

tial

of S

WH

S (s

q.m

.) Total Technical Potential = 353300 sq. m.

Economic potential (limit=payback period of 5 years) =19700 m2 collector area

Framework for Potential Estimation of Solar Water Heating Systems in a Country

Country details(Area, Average Weather

Data)Locations where

weather data available

Locations where weather data unavailable

Selection of base city

Methodology for potential estimation for a target

area

Weather data

End use details for each sub-class

Identification of sectors and classification within each

sector

Potential of SWHS in base city

Potential of SWHS in different location

Identification of variables for a different

location

Weather data

End use detailsSpatial

Interpolation

Potential of SWHS in nearby area where weather

data is not available locationAggregation for all the

locationsPotential of SWHS in the

country Technical potential Electricity savings

Diffusion of SWH

0

50

100

150

200

250

300

1990 2010 2030 2050 2070 2090

Year

Sola

r W

ater

Hea

ting

Cap

acit

y (c

olle

ctor

are

a in

mil

lion

sq

. m.).

.

Actual installed (million sq. m.)Potential 140 million sq. m.Potential 60 million sq. m.Potential 200 million sq. m.Extrapolated Potential (million sq.m.)

Potential = 60 million m2

Potential = 140 million m2

Potential = 200 million m2

Estimated Potential in

2092 = 199 million m2

Load curve of a typical day –MSEB (8/11/2000 source: WREB annual report-2001)

10260 MW9892 MW

6000

7000

8000

9000

10000

11000

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

De

ma

nd

, MW

morning peak

Evening peak

Sample Industrial Load Profile (Mumbai)

Time of Use Tariff (MSEB-HT Ind., Jan

2002)

0

50

100

150

200

250

300

350

400

450

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

Pai

se

/kW

h

Off-peak

Peak

Partial Peak

Peak

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

30 T MeltingArc furnace

Bar mill

Wire mill

40 T Melting Arc furnace

St. steel Scrap mix orAlloy steel scrap mix

Alloy steel scrap mix

Convertor (only for St Steel)

Ladle Arc furnace

VD or VOD station

Bloom caster

Billet caster

Bloom mill

oooooo

Reheat furnace

Reheat furnace

Reheat furnace

Wire products for final finish

Rods, Bars for final finish

Open store

Open store

Open store

Open store

Steel Plant Flow Diagram

Flour Mill

0

10

20

30

40

50

60

Time hours

Lo

ad M

W

Optimal with TOU tariff

Optimal with flat tariff

2 4 6 8 10 12 14 16 18 20 22 24

Steel Plant Optimal Response to TOU tariff

Process Scheduling Summary

Example Structure Results SavingFlour MillContinuous

Linear, IP120 variables46 constraints

Flat- 2 shift - 25%storeTOU-3 shift

1% 6.4%75%peakreduction

Mini Steel PlantBatch

Linear, IP432 variables630 constraints

FlatTOUDiff loading

8%10%50% peak reduction

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

Cool Storage of Commercial Complex -under TOU tariff

129 kVA

208 kVA

0

50

100

150

200

250

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

kVA

with optimal cool storage

Load following (without storage)

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.

Willans Line

LP Steam 5. 5 b, 180 oC

Gas turbine -1

Boiler

ST

PRDS-1

PRDS-3

Condenser

Deaerator

Process Load

Process Load 40 T/h

G

1

G

4

Process Load, 60 MW

BUS

Grid 7.52 MW

Feed water426.5 T/h

SHP Steam 100 bar,500o C

HP Steam 41b,400 oC

Fuel, LSHS9.64 T/h

Fuel, HSD 5.9 T/h

WHRB-1

Supp. FiringLSHS 5.6 T/h

Stack

20 MW

Process Load,125 T/h

Process Load,150 T/h

MP Steam 20b, 300 oC

PRDS-2

Gas turbine -2

G

1

WHRB-2

Supp. FiringLSHS 5.6 T/h

20 MW

Fuel, HSD5.9 T/h

136 T/h

136 T/h

131.7 T/h12.5 MW

76.2 T/h60.6 T/h

117.1 T/h

40 T/h 49.5 T/h 16.2 T/h

20 T/h

40 T/h

53.4 T/h

Make up water,357 T/h

Cogeneration Example

Import Power from Grid with Cogeneration for a Petrochemical Plant

11 MW

17.621.6

00

5

10

15

20

25

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

Impo

rt po

wer M

W

flat tariff TOU tariff

peak period demand

Export power to the grid with Cogeneration for a Petrochemical Plant

0

10

20

30

40

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

Expo

rt Po

wer M

W

flat tariff TOU tariff

9.7 MW

Peak perioddemand

- Integrated approach

Industrial Load Management

Operating cost structure

Optimal operating strategy of captive/cogeneration plant

Captive/Cogeneration power model

Grid tariff, fuel costs, Grid conditions

Modified process demand profile

Process demand profile, Cooling electric load profile, Steam load profile

Process load model Air conditioning

(cooling) load model

Optimal process load schedule Optimal cool storage

Plant/measured input data

Modified cooling electric load profile

Modified steam load profile for process related loads

Glass furnace Classification of furnace

Type of firing (Cross fired / end fired)

Raw material Batch material (like silica,

soda ash etc.) Cullet (recycled glass)

Heat source Flame direct contact with

glass Minimum energy

requirement Heating of raw material up

to reaction temperature Endothermic heat of

reaction for batch material

Doghouse (raw material feeding section)

Throat (processed glass outlet)

Melting end

Regenerator

Checker work

Working end

Modeling practices for glass furnace Continuum Process

model Commonly used Glass furnace process

in continuum equation Three dimensional

Navier-Stokes Equation and Hottel’s zone method for radiation

Process models used mainly troubleshooting and screening variables

Limitations of process models

Data intensive inputs Needs specialized skills

and computational facilities to use

Energy performance not studied

Not easy to link operating parameters and impact on energy performance

Approach for study

Conducting measurements for operating parameters not captured in existing instrumentation

Overall energy and mass balance

Study of operating glass furnaces

Identifying key operating variables

Analyzing time series data of key operating variables present in existing instrumentation

Literature search for furnace modelling

Refining assumptions and empirical relationships with experimental measurements

Developing mathematical furnace models with simplified assumptions for sub-processes

Solving these models for operating variables

Coupling models for understanding overall performance

Establishing relationship between dependent and independent variables empirically and analytically

Comparing measured parameters and model result

Conducting parametric of variables using validated models

Identifying areas for energy performance improvement and optimal operating strategy

Control volume

Combustion Space

Molten glass

Fuel

Batch Glass

Regenerator

Exhaust Gas

Combustion Air

Control Volume 1

Control Volume 2

Control Volume 3

, , , , ,,100

, , ,

( )obh bh g l wall g g g g rk bh f bh f w w latw C

g sensi g rk bh f w

m h Q Q m h m h m h m h h

Q Q Q Q

, , , , , , , , , , , ,

, , ,,

fu comb air nonreg noncomb air nonreg air air comb reg air comb reg l wall comb g tot f f

fu air reg l reg fair nonreg

m CV m m h m h Q Q m h

Q Q QQ

, , , , , , , , , , , , , , , , , , ,f tot in f in air leak reg air leak f tot out f out l wall reg air comb reg air comb reg out air comb reg inm h m h m h Q m h h

Eq. 1

Eq. 2

Eq. 3

Mass balance of furnace

Input streams Batch material

Cullet (recycled glass)

Raw material Moisture

Fuel Combustion air (from

regenerator) Air leakage (Any air

other than inlet from regenerator)

Output streams Molten glass

Cullet Glass from raw material

Flue gas to regenerator Combustion products Glass reaction products Water vapors Air (Not reacted in

combustion) Flue gas leakage from

furnace

Mass balance estimationEstimation of flue gas

formation

Based on stoichiometric Calculation of combustion

Products of combustion

Air leakage

No methodology for estimation in literature

Moisture in batch

Based on % in batch

Products of glass reaction

Based on stoichiometric Calculation of glass

Species in furnace flue gas

CO2

H2O

SO2

O2

N2

Oxygen % in flue gas (v/v dry basis)

Used as indicator for excess air control

Air leakage estimation Furnace operates

positive pressure Air leakage in local

negative pressure area

Air leakage due to higher pressure on air side

Air supplied for atomization and flame length control

Air for fuel atomization / flame control during firing and tip cooling air during non firing

Air induced by jet effect of burner

Combustion air from regenerator

Air leakage from furnace joints

Air leakage from flux line cooling

Glass melt

Energy balance for furnace Input streams

Energy from fuel Energy from

preheated combustion air

Energy from batch material

Energy from air leakage

Output streams Energy carried in glass

Heat of reaction Sensible heat of glass

Energy carried in flue gas Energy for air leakage Energy for batch gases Energy for moisture Energy for combustion

air Energy loss from walls

Surface heat loss from walls

Radiation losses (due to opening)

Energy balance glass melt Heat of reaction for glass Heat carried by glass Heat carried by batch gas

Heat carried away by glass

Heat carried by batch gases and moisture

Endothermic heat of reaction for glass formation

Furnace wall losses

Molten Glass

Glass flow direction

Flux lineZones along furnace sidewall depth

Zones along furnace melter sidewall length

Zones along furnace crown and superstructure side wall length

Furnace model input parameters

Design parameter Design capacity of

furnace Melting area Length to width ratio Height of combustion

volume Refractory and

insulation details

Operating parameters Furnace draw Type of fuel Batch to cullet ratio Moisture in batch Furnace pressure Oxygen at furnace

outlet Atomization pressure Reversal time Flux-line and burner tip

cooling air pressure

Model flow diagram

Mass of air

Flue gas leakage

Oxygen % at regenerator outlet

Design

variables

Guess for total heat added

Fuel stoichiometric calculation

Glass reaction calculation

Furnace air / flue gas leakage calculations

Gap in flux line Gap near burner

Furnace operating pressureCooling air velocityNumber of burner

Burner air nozzle diameter

Furnace design capacity

Melting area

Furnace design details

Color of glass

Furnace geometry

Air leakage

Regenerator calculation

Flue gas outlet temperature

Heat loss from flue gas

Heat loss from regenerator wall

Oxygen % at furnace outlet

Combustion zone stoichiometric calculation

Furnace wall lossesFurnace operating

characteristics

Heat of reaction and heat carried by glass

Mass of flue gas

Heat loss from furnace area wall

Gas from glass reaction

Raw material compositionFurnace geometry calculation

Furnace design characteristics

Heat carried with glass

Heat of reaction for glass

Heat loss batch gas

Heat loss from batch moisture

Total heat added in furnace

Fuel calculationFuel calorific value

Fuel composition

Glass composition

Moisture in batch and cullet

Cullet %

Glass draw

Fuel consumptionCombustion species

Heat loss from flue gas leakage

Heat loss from air leakage

Ambient conditions

Glass outlet temperature

Port neck

Checkers packing

Glass level

1

2

5

Manual damper for airflow selection and control

6

7

Diverter damper

3

4

8

Measurement locations Combustion air

Furnace measurement

Measurement location

Type of measurement

1

Oxygen % , Pyrometer checkers surface temperature

2Oxygen %, Flue gas temperature

3Oxygen %, Flue gas temperature

4Oxygen %, Skin temperature

5Pyrometer checkers surface temperature

6Velocity of air at the suction of blower

7Outside wall temperature for crown and side wall

8Pyrometer glass surface temperature

Model results: Actual SEC

2.8%(118)

0.7%(30)69

1%(45)

9.7%(414)

38.2 % (1628)

2%(84)6.1%

(261)5% (212)

4.6% (198)

29.4%(1256)

33.8%(1485)

69% (2939)

Heat carried in glass

Furnace wall losses

Heat lost in moistureHeat of glass

reactionBatch gas losses

Heat loss from furnace opening

Heat lost steel superstructure

Regenerator wall losses

Heat loss from flue gas

Heat lost in cold air ingress

Heat recovery in air heating

100%(4267)

Energy introduced in furnace

From fuel 134% (5752)

Heat carried in regenerator from flue gas

Model results: Target SEC

1.7 % (63)

1.2% (45)

10.5% (390

42.7 % (1628)

1.6 % (60)7 %

(262)5.3 % (196)

5.6 % (211)

23.5 % (876)

40.5% (1510)

69.6 % (2597)

Heat carried in glass

Furnace wall losses

Heat lost in moistureHeat of glass

reactionBatch gas losses

Heat loss from furnace opening

Heat lost steel superstructure

Regenerator wall losses

Heat loss from flue gas

Heat recovery in air heating

100 % (3730)

Energy introduced in furnace

140 % (5240)

Heat carried in regenerator from flue gas

Conclusions Target SEC estimated for

16 industrial furnaces Effect of furnace draw on

target SEC is demonstrated

0

2000

4000

6000

8000

10000

12000

14000

16000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Furnace number

SE

C (

kJ

/kg

)

Target SEC Actual SEC

0

2000

4000

6000

8000

10000

12000

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280

Draw (TPD)

Tar

get

SE

C (

kJ/k

g)

Generalized approach for model based benchmarking

Survey of existing models of process Developing

experimentation protocols

Study of actual process operation (process audit)

Operating procedure and practices

Control strategy and instrumentation

Process constraints Logbook parameters

Understanding basics Defining system

boundary Writing fundamental

equations governing process

Decide assumptions Identifying empirical correlations for process

Model development Divide process into sub-

models Identify input / output

parameters for sub-models Identification of design and

operating variables Developing linkage between

process parameters and energy consumption

Experimentation

Validation of model

Refinement of model

Data from industrial process

Usage of model

Target energy estimation

Parametric analysis

Energy intensive process

Energy Planning Research Objective

Energy Planning - Aggregated at National Level

Sectoral Planning - Oil, Electricity Limited Efforts in Local Level Energy

Planning Energy critical input for development Need for micro level energy planning Objective: Develop a tool for improving

local level energy decisions

DSS Modules Energy Models in Literature – aggregate ,

homogenous unit, optimisation, no linkage with decision structure

Study of decision structure, proposed an accounting framework

Disaggregation by sector, end-uses DST-UNDP project – Bankura district in West

BengalAnalysis District level – with block as an unit

Block level- with village as an unit

Energy decisions for district officials

District Magistrate Ratifying departmental decisions

Zilla Parishad Sabadhipati Sanctioning/selecting schemes

Manager, DIC Sanctioning industries, biogas plant, Priority list of industries.

DE,WBSEB/Addnl CE(Rural Elec)

Mouza electrification

Controller food,civil supplies Distribution of kerosene, coal.

Lead Bank Officer Preparation of District credit plan

District forest Officer (DFO, working plan)

Managing forest area

District Planning Officer Prepares district plan

Typical Energy Decisions District – Fund Allocation to blocks, Mouza

electrification, Industrial devpt., Coal – elect., fuel / ration shops Sanctions.

Block– Fund Allocation to GPs, Kerosene allocation, industry promotion, marketing support.

Gram Panchayat – Agriculture / irrigation schemes, Co-op industry, request for fuel/ration shop, electricity.

Household – Fuel choice, Device choice.

60

SECONDARY

DATA

REMOTE SENSED DATA

PRIMARY DATA

DIGITISED MAPS

D A T A B A S E

Identify indicators/variables

affecting energy

Trends in indicators

ENERGY DSS

SUPPLY MODULE

DEMAND MODULE

FUTURE ENERGY DEMANDS BY

SECTOR /END USE

FEASIBLE ENERGY SUPPLY SCENARIOS

IMPACT ASSESSMENT / EVALUATION

DEVELOPMENT PROFILE

DSS Frame Work

Ground Truthing

Digitized Maps

Map Showing Landuse

Classification Areal

extent of forest, non-

forest

Remote Sensed images

Sub Classification of Sal, and Mixed

forests

GIS ANALYSIS

Secondary Data

Estimation of wood volume correlations

Height, Density measurements

Standing wood estimates forest

areas

% of Fuelwood

Fuelwood avail from

forest areas

Fuelwood non-forest

Total Fuelwood available

Average fuelwood density non-forest

(Survey / literature)

Estimation of crop areas

Total Crop residues available

Crop residue factors

CROP RESIDUE MODULE

ACCOUNTING MODULE

ACCOUNTING MODULE

FUELWOOD MODULE

Drivers for Demand Scenarios

Residential cooking

Residential Non-cooking

Agriculture Industrial

Drivers Population Low-Med-High(L-M-H)

IncomeNo Transition(NT)/ Transition(T)

Population (no of Households) Low-Med-High (L-M-H)

Income NT-T

Electrification

NE-ME-AE

Rainfall

Low-Med-HighLR – MR – HR Irrigation

NI – LI- MI – HI

Crop Pattern NTC/TC1/TC2 Pump Electric NE/ME/AE

Growth

Low-Med-HighLG/MG/HG

*Rice mills linked to agricultural scenarios

No of combinations

6 18 108 3+

Population

Income Village Electrification

Rain Irrigation Land

Crop Pattern

Pump Electrified

Ind. Growth

SC1 L NT NE HR NI NTC NE LG

SC2 H T AE LR HI TC2 AE HG

SC3 M T ME MR MI TC2 ME MG

SC4 L T AE HR HI TC1TC2

AE HG

SC5 L T ME HR MI TC1 ME MG

Select Scenario cases

Electricity Demand in 2005

154

328283

557

421

0

100

200

300

400

500

600

SC1 SC2 SC3 SC4 SC5

GW

h in

200

5

136 GWh

Number of Electric Pumpsets

Usage and seasonal variation

Technical specifications

Agriculture Load

Number of Commercial shops

Technical Characteristics and Usage

Total Appliances

Appliance ownership

Commercial Load

Total Electricity demand. Daily load Curve Summer/Winter

Appliance ownership data

by each income class

Total no of

Appliances in Area

Electricity consumption in

Residential sector. (Including

the load curve) for each

season/Annual

Demographics

1. Number of Households (HH)

2. % of HH in each income class

Technology Characteristics 1. Rating (kW) 2. Efficiency

Usage pattern 1. Daily variation 2. Seasonal Variation

Village Electrification example

Below Poverty

33%Middle Income

23%

High Income13%

Above Poverty

31%

Income class distribution in Rajamele in 1994.

0

10

20

30

40

50

60

70

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

Hour of the day.

Load

(kW

)

Residential Electricity Dem and Agriculture Dem and

Industry dem and Total Load

Daily load curve for Rajamele village in Summer in 2005

DSS Capability/Use Energy Quantification- from data already

available/collected by Govt Assess impacts of development paths Quantify future commercial fuel demands –

elec/coal/LPG/kerosene Rural Electrification- Load Profile

Estimation,Sizing, Option Selection, Impact Assessment

Spatial Representation of results Classification of areas as biomass surplus/deficit

Summing Up Examples – illustrate variety of

optimisation/ simulation models for energy sector

Decision context and model formulation critical

Reality check - Applicability Generalisation important but.. Data intensity and uncertainty

Acknowledgment

Balkrishna SurveProject Assistant

S. AshokPh.D - 2003

Thank you

Sandeep Phulluke, Sagar Kasar, Abhijit Bhure,Amit Khare, Jagdish

DSS Project Staff(2004-2006)

Santanu B. Faculty

Indu PillaiPh.D - 2008

Vishal S. Ph.D. - 2008

Arun P. Ph.D. - 2009

U.N. GaitondeFaculty

Manojkumar M.V.M.Tech - Ongoing

Mel George A.M.Tech - Ongoing

References Ashok.S and R. Banerjee , “An Optimisation model for Industrial Load management”, IEEE Trans on Power

Systems, Vol.16, No. 4, Nov.2001, pp 879-884. Ashok S and R. Banerjee , Optimal Operation of Industrial Cogeneration for Load Management, IEEE

Trans on Power Systems, Vol. 18, No. 2, MAY 2003. Ashok.S and R.Banerjee, Optimal cool storage capacity for load management , Energy, Vol. 28, pp 115-

126, 2003. Vishal S, U.N.Gaitonde,R Banerjee., “Model based energy benchmarking for glass furnace”, Energy

Conversion and Management, Vol.48, pp 2718-2738, 2007. R. Banerjee, A. B. Inamdar, S. Phulluke, B. Pateriya, “Decision Support System for Energy Planning in a

District”, Economic and Political Weekly, Vol. 34, No. 50, pp 3545-3552, 1999. R.Banerjee, Rahul Pandey, Abhijit Bhure and Sandeep Phulluke ,“Electricity Demand Estimation for Village

Electrification ”, Proceedings of National Renewable Energy Conference , IIT Bombay, Nov. 30-Dec.2 ,2000.

Rahul Pandey, R.Banerjee, Abhijit Bhure and Sandeep Phulluke ,“Framework for Design and Evaluation of Electric Supply Options for a rural area”, Proceedings of National Renewable Energy Conference , IIT Bombay, Nov. 30-Dec.2 ,2000.

Arun P., Santanu Bandyopadhyay and R. Banerjee, ‘Sizing curve for design of isolated power systems’, Energy for Sustainable Development,Volume XI, No. 4, December 2007.

Indu R. Pillai and R. Banerjee, ‘Methodology for estimation of potential for solar water heating in a target area’, Solar Energy, Vol.8, No.2, pp 162-17, 2007.

George, R. Banerjee,Analysis of impacts of wind integration in the Tamil Nadu grid, in press, Energy Policy UK Wind speed data: GWEFR Cyf Hourly-mean wind speed datasets for sites in the European

Wind Atlas, available at http://www.gwefr.co.uk/datasets.htm UK load curves: Demand Data, UK National Grid, available at

http://www.nationalgrid.com/uk/Electricity/Data/