Strategies to reduce transmission costs for solar …...Strategies to reduce transmission costs for solar plants in the National Electricity Market Ben Elliston and Iain MacGill School

Strategies to reduce transmission costs for solar plants in the National

Electricity Market

Ben Elliston and Iain MacGill

School of Electrical Engineering & TelecommunicationsCentre for Energy and Environmental Markets

UNSW

b.elliston@student.unsw.edu.au

In the news

An aerial impression of the Kogan Creek Solar Boost Project. Source: CS Energy

Outline

● Brief overview of plant technologies● Transmission investment goals● Transmission in the NEM● Solar plant output simulations● Analysis of results● Conclusions

Ground mounted PV plants

34MW PV plant in Arnedo, La Rioja, Spain. Source: Renewable Energy magazine

Two-axis tracking PV plants

1MW PV plant in Madera, California. Source: PROINSO

Concentrating Solar Power (CSP)

SEGS plant in Kramer Junction, California. Source: TREC­UK

Concentrating Solar Power (CSP)

Andasol­1 plant, Granada, Spain (7.5 hours TES). Source: German Aerospace Center

Transmission investment goals

● From Review of Energy Market Frameworks:– Timely processing of connection applications

– Cost-reflective pricing

– Efficient level of investment

● Main issues for renewable energy– Build in high quality locations (further from grid),

or lower quality locations (closer to grid)?

– Plants built faster than transmission2-3 years vs 8-12 years (Mills et al., Cavallo)

Transmission in the NEM

NEM regions (interconnectors in red)

Transmission in the NEM

● Open access to transmission network● Costs are recovered from loads, not generators

– “Under-signalling of network costs for new generators” (Frontier Economics)

● Shallow connection charging● Augmentation subject to RIT-T

– Net benefit (there may be cheaper measures)

– Evidence of intraregional mis-pricing

– May be slow: EIS, land title, build-out (8-12 yrs)

Rule change: Scale Efficient Network Extensions (June 2011)

● Proposed by MCE● Encourage efficient network development

– Identify regions with high renewable potential

– Forecast future generation

● A chicken and egg problem– One generator ready now, but more expected

● Risks for generators, TNSPs

– Risks building transmission to right scale● Stranded assets● Inefficient duplication

Rule change: Scale Efficient Network Extensions

● Generators are given option to connect● Once first generator connects, build● Shallow connection from generators to new

network nodes

Under current arrangements, generators ..

● may cluster around existing transmission● will minimise cost of grid connection● may form agreements with nearby loads ● may encounter constrained transmission

Transmission sizing strategies

● High capacity factor generators(e.g. coal-fired plants meeting base load)

– Size transmission line to rated output

– “Get all power to market”

● Low capacity factor generators(e.g. CCGT meeting peak load)

– Size transmission line to rated output

– “Get high value power to market”

● Solar generators?

Prior work

● Work on sizing in the wind sector– Cavallo (1995) – oversizing farms

– Boerema and MacGill (2010)

– Pattanariyankool & Lave (2010)

● Recent work on storage in CSP systems– Johnston (2009): CSP in the NEM

– Beyond Zero Emissions ZCA Plan (2010)

– Denholm, et al (2010): role of storage

– Wittmann, et al (2011): operation strategies

Solar plant output modelling

● System Advisor Model (NREL)● Modelled 100MW plants in Cobar, NSW

Source: Google Maps

Solar plant output modelling

● System Advisor Model (NREL)● Modelled 100MW plants in Cobar, NSW● 2009 weather data● Modelled systems

– PV

– 2-axis tracking PV

– CSP (w/ thermal storage)

● NEM spot price data (NSW, 2009) from AEMO– Half hourly, averaged to hourly

PV output (Jan. 1 & 2, 2009)

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

Po

we

r (k

W)

PV output (every noon, 2009)

0

10000

20000

30000

40000

50000

60000

70000

80000

90000kW

PV results

Transmissioncapacity (MW)

Annual yield (GWh)

Energy fraction

Revenue ($)

Revenue fraction

90 178.31 1.000 6,601,000 1.000

80 178.17 0.999 6,596,000 0.999

70 174.74 0.980 6,443,000 0.976

60 164.99 0.925 6,029,000 0.931

50 148.93 0.835 5,418,000 0.821

PV and 2-axis tracking PV output(Jan. 1 & 2, 2009)

-20000

0

20000

40000

60000

80000

100000

PV2-axis PV

Po

we

r (k

W)

2-axis tracking PV results

Transmissioncapacity (MW)

Annual yield (GWh)

Energy fraction

Revenue fraction

90 251.70 1.00 1.00

80 242.71 0.96 0.97

70 223.58 0.89 0.89

60 198.07 0.79 0.79

50 168.98 0.67 0.67

Trough net outputJan. 1 & 2, 2009

-20000

0

20000

40000

60000

80000

100000

120000

PVCSP-2.5

Ne

t Po

we

r (k

W)

Trough net outputJan. 1 & 2, 2009

-20000

0

20000

40000

60000

80000

100000

120000

PVCSP-2.52h

Ne

t Po

we

r (k

W)

Trough net outputJan. 1 & 2, 2009

-20000

0

20000

40000

60000

80000

100000

120000

PVCSP-2.52h4h

Ne

t Po

we

r (k

W)

Trough net outputJan. 1 & 2, 2009

-20000

0

20000

40000

60000

80000

100000

120000

PVCSP-2.52h4h6hN

et P

ow

er

(kW

)

Trough plant energy yield

0.5 0.6 0.7 0.8 0.9 1.0

0.00

0.20

0.40

0.60

0.80

1.00

1.20

0h TES2h TES4h TES6h TESPV

Constraint fraction

En

erg

y fr

act

ion

Trough plant revenue

0.5 0.6 0.7 0.8 0.9

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

0h TES2h TES4h TES6h TESPV

Constraint fraction

Re

ven

ue

fra

cti o

n

“Ideal” generation curve

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

0

20000

40000

60000

80000

100000

120000

PV2-axis PVIdeal (no storage)

Hour

Ne

t po

we

r (k

W)

“Ideal” generation curve

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

0

20000

40000

60000

80000

100000

120000

PV2-axis PVIdeal (no storage)Ideal (some storage)

Hour

Ne

t po

we

r (k

W)

“Ideal” generation curve

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

0

20000

40000

60000

80000

100000

120000

PV2-axis PVIdeal (no storage)Ideal (some storage)Ideal (more storage)

Hour

Ne

t po

we

r (k

W)

“Ideal” generation curve

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

0

20000

40000

60000

80000

100000

120000

PV2-axis PVIdeal (no storage)Ideal (some storage)Ideal (more storage)Ideal (even more storage)

Hour

Ne

t po

we

r (k

W)

“Ideal” generation curve

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

0

20000

40000

60000

80000

100000

120000

PV2-axis PVIdeal (no storage)Ideal (some storage)Ideal (more storage)Ideal (even more storage)

Hour

Ne

t po

we

r (k

W)

80MW limit~20% less energy

Conclusions

● In restructured industries (NEM), transmission investment needs to be carefully directed

● Risk, not cost, the biggest problem● Modular construction a big advantage● Over-sizing fixed PV systems very effective● More CSP storage → higher impact of

constraint (just like fossil plants)● Better to match power block to available

transmission (less $, run at full load)

Questions?

Ben Elliston and Iain MacGill

School of Electrical Engineering & TelecommunicationsCentre for Energy and Environmental Markets

UNSW

b.elliston@student.unsw.edu.au