Transcript Integrating Renewables into the Electricity System Overview

Integrating Renewables into
the Electricity System
An Historical Overview
Professor Michael Laughton
Centre for Energy Policy & Technology,
Imperial College
Open University Jan 24th 2006
Renewable Energy Sources
Marine
Wave
Tidal
Hydro
Large-scale, Small-scale
Wind
Onshore, Offshore
Solar
Passive-, Active-heating,
Photovoltaic
Hot-dry rocks, aquifers
Geothermal
Biofuels
(onshore, offshore)
(barrage, stream)
Waste, Crops, Landfill gas
(combustion, conversion)
Characteristics of variability
Fluctuations and possible power supply shortages
due to:
• Uncertainties in prediction of occurrence
• Power conversion plant limits
- too little resource availability
- too much resource availability
• Magnitude of fluctuations
- small
- large
• Speed of fluctuations
- slow (usually predictable)
- fast (less predictable)
Examples are as follows: -
Typical annual variation in wave power
levels
Wind Turbine Output Characteristic
% Rated Output
120
100
80
60
40
20
0
0
5
10
15
20
Wind Speed
m/s
25
30
Slow strong fluctuations
Reference: EON_Netz_Windreport_e_eng.pdf
Fairly rapid decrease
Winter power infeed E.ON control area 17.11 to 23.11.03
Reference: EON_Netz_Windreport_e_eng.pdf
Fast fluctuations
Euros / MWh
Danish electricity spot prices first week of January
2005
Met Office data for wind speeds
http://www.metoffice.com/education/
archive/uk/
Note: 9 knots = 4.63 m/s
(Wind Turbine cut-in speed approx)
Note, however,……..
• Many Met Office stations are
geographically irrelevant for judging wind
power potential in Britain.
• Wind speeds at Met Office station
monitoring heights need to be increased to
account for variation of speed with height.
• A very simple rule might be
Vz = Vh (z / h)a
where Vz and Vh are wind speeds at
heights z and h, h>z and a = 0.16
Ref: Halliday and Lipman, 1982
“Economic and Operational Assessment
of Intermittent Generation Sources on
Power Systems”
Colloquium EE Dept, Imperial College, 5 March 1987
Contributions from:
E.D.Farmer (Imperial College from CEGB)
D.J.Milborrow (CEGB)
J.P Palutikof, C.P.Watkins (UEA)
S.C.Ryrie (Bristol Poly)
D.T.Swifthook (CEGB)
P.R.Hanson (CEGB)
M.J.Grubb (Imperial College)
A.Thorpe (CEGB)
D.G.Infield, J.A.Halliday (Rutherford-Appleton Laboratory)
Cubed values of annual mean wind speeds at Southport
Marshside proportional to wind turbine power output
Ref: J.P.Palutikof, C.P.Watkins, “Some Aspects of Wind speed Variability….”, Op. Cit
Probabilistic electricity generation analysis
is needed to determine capacity credit
• ONLY direct time series analysis of historical data of
wind combined with probabilistic analysis of the
availability of thermal units can hope to capture the
real capacity credit of wind.
• The risk of system failure within a few GW of peak
demand is not much less than at peak demand,
• BECAUSE the thermal plant output may have a
standard deviation of between1 and 2 GW.
Ref: M.J.Grubb, “Capital Effects at Intermediate and Higher Penetrations”, Op Cit
Probabilistic electricity generation analysis
is needed to determine capacity credit
• Small capacity shortages have a much higher
probability than large shortages and have little effect
on security of supply,
• BUT as the capacity of wind in the system increases,
the capacity credit is increasingly dominated by the
smaller likelihood of little or no output.
Ref: M.J.Grubb, “Capital Effects at Intermediate and Higher Penetrations”, Op
Cit
Baseload capacity displacement with
increasing wind penetration
Variations
with peak
availability,
diversity,
system
limiting costs
Conclusion:As a ‘Rule of
Thumb’ the
capacity credit
for wind in
Britain is the
square root of
the GW of
wind installed
Ref: M.J.Grubb, “Capital Effects at Intermediate and Higher Penetrations”, Op Cit
Percentage of time over a 5 Year Period
Total wind power generation distribution
to achieve half Government 2010 target
800
Average hourly generated power MW
Source: National Grid PIU Supplementary Submission 28 Sept 02
TM / ML / 03-04-02
Comment
• The above slide is key to understanding wind
capacity credit.
• It is based on a study by National Grid of ten years of
hourly Met Office data for sites relevant to the mainland
Britain power transmission system.
• The graph shows the probability of actual wind power
generated per annum from 7600 MW of installed
capacity assuming no transmission constraints.
Further National Grid studies
• Purpose - to establish reliability of supply with increasing wind
penetration.
• The following charts show the probability density distributions of the
availability of the extra capacity needed to maintain security of supply
levels.
• The first chart relates to the conventional thermal plant planning margin
of 19% calculated from plant availability statistics.
• The third chart shows the influence of the graph shown on the previous
slide relating probabilities of wind power output to installed wind capacity
in a combined probabilistic analysis of existing thermal plant and wind
capacity.
• This next slide and also the slide following draw attention to the
implications for capacity credits of wind and the need to maintain
conventional plant capacity. The results are similar to those obtained by
M.Grubb in the 1980’s.
Total generation capacity for secure supply
Zero indicates generation balances load. Area to left of zero is the probability of
not meeting 50,000 MW peak demand 10 winters per century
30,000
25,000
20,000
15,000
10,000
5,000
0
-5,000
-10,000
25,000 MW wind
55,000 MW conventional
Spare capacity = 30GW
Source:
NGC
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
-2,000
-4,000
-6,000
7,500 MW wind
57,000 MW conventional
Spare capacity = 14.5GW
-8,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
-2,000
-4,000
-6,000
-8,000
500 MW wind
59,000 MW conventional
Spare capacity = 9.5GW
National Grid generation capacity
calculation to maintain security of supply
standards – a different presentation
20,500 MW
extra
Plant Capacities MW
90000
Plant Capacities
80000
70000
25,000 MW Thermal
60000
40000
Thermal planning
margin reduced by
4,500 MW
30000
25,000 MW Wind
20000
25,000 MW Thermal
50000
10000
0
1
2
Wind Penetration
3
The extra system
support costs relate
to the net 20,500
MW of thermal
capacity not
displaced
Note: The thermal capacity not displaced has also been called standby or shadow capacity
Further results from the ILEX study for the DTI on
system costs of additional renewables
(the SCAR Report
•
•
•
•
In this study equivalent thermal capacity was removed against the energy
contributions from increasing wind capacity assuming no transmission
limitations exist.
A combined probabilistic simulation of operation then established the levels of
extra standby plant needed. This is a capacity remix that was not present in the
studies of Grubb and the National Grid shown above.
Overall the results confirm that large amounts of wind power need large
amounts of conventional plant to be retained. It is not clear what to call this
retained capacity. ‘Standby capacity’ has been used by the Royal Academy of
Engineering and its equivalent replacement costed accordingly. The German
utility E.ON Netz refer to it as ‘shadow capacity’. Obviously the ‘standby costs’
would be very different from those quoted in the SCAR Report.
The astonishing conclusion from all of these studies (Grubb, National
Grid, ILEX) is that regardless of the wind capacity in the system, the
conventional capacity needed always exceeds the peak demand.
Demand growth scenarios with various
penetration levels of wind energy by 2020
Peak demand 75,700 MW; Other renewables 1,600 MW
Total
Installed
wind
wind
energy capacity
%
MW
Conventional Conventional
capacity
capacity
required
margin
MW
%
Spare
capacity
margin
MW
Spare
capacity
margin
%
0
0
90,083
19
15,983
21
10
9,900
86,800
15
22,600
30
20
24,000
84,000
11
33,900
45
30
38,000
82,500
46,400
61
9
Ref: “Quantifying the System Costs of Additional Renewables in 2020”, ILEX Energy
Consulting Report to the DTI, October 2002.
Conclusions - Integrating Renewables
into the Electricity System
Need to know more about the interaction of
• rates of change, magnitudes and lengths of
intermittency with conventional plant needs,
• system constraints (loadflow / transmission
constraints, voltage and frequency control),
• effects of groupings of wind generation with
regard to capacity credit,
• how to define such groups (size, spread,
location, geographical orientation,….),
• etc, etc…..