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