Climate change

September 21, 2009

Solar realities and transmission costs – addendum

Peter Lang’s ’solar realities’ paper and its associated discussion thread has generated an enormous amount of interest on BraveNewClimate (435 comments to date). Peter and I have greatly appreciated the feedback (although not always agreed with the critiques!), and this has led Peter to prepare: (a) an updated version of ‘Solar Realites’ (download the updated v2 PDF here) and (b) a response paper (download PDF here). Below I reproduce the response, and also include Peter’s sketched analysis of the scale/cost of the electricity transmission infrastructure (PDF here).

———————————————–

Comparison of capital cost of nuclear and solar power

By Peter Lang (Peter is a retired geologist and engineer with 40 years experience on a wide range of energy projects throughout the world, including managing energy R&D and providing policy advice for government and opposition. His experience includes: coal, oil, gas, hydro, geothermal, nuclear power plants, nuclear waste disposal, and a wide range of energy end use management projects)

Introduction

This paper compares the capital cost of three electricity generation technologies based on a simple analysis. The comparison is on the basis that the technologies can supply the National Electricity Market (NEM) demand without fossil fuel back up. The NEM demand in winter 2007 was:

20 GW base load power;

33 GW peak power (at 6:30 pm); and

25 GW average power.

600 GWh energy per day (450 GWh between 3 pm and 9 am)

The three technologies compared are:

1. Nuclear power;

2. Solar photo-voltaic with energy storage; and

3. Solar thermal with energy storage

(Solar thermal technologies that can meet this demand do not exist yet. Solar thermal is still in the early stages of development and demonstration. On the technology life cycle Solar Thermal is before “Bleeding edge” – refer: http://en.wikipedia.org/wiki/Technology_lifecycle).

This paper is an extension of the paper “Solar Power Realities” . That paper provides information that is essential for understanding this paper. The estimates are ‘ball-park’ and intended to provide a ranking of the technologies rather than exact costs. The estimates should be considered as +/- 50%.

Nuclear Power

25 GW @ $4 billion /GW = $100 billion (The settled-down-cost of nuclear may be 25% to 50% of this figure if we reach consensus that we need to cut emissions from electricity to near zero as quickly as practicable.)

8 GW pumped hydro storage @ $2.5 billion /GW = $20 billion

Total capital cost = $120 billion

Australia already has about 2 GW of pumped-hydro storage so we would need an additional 6 GW to meet this requirement. If sufficient pumped hydro storage sites are not available we can use an additional 8GW of nuclear or chemical storage (e.g. Sodium Sulphur batteries). The additional 8 GW of nuclear would increase the cost by $12 billion to $132 billion (the cost of extra 8 GW nuclear less the cost of 8 GW of pumped hydro storage; i.e. $32 billion – $20 billion).

Solar Photo-Voltaic (PV)

From ‘Solar Power Realities’ :

Capital cost of PV system with 30 days of pumped-hydro storage = $2,800 billion. (In reality, we do not have sites available for even 1 day of pumped hydro storage.)

Capital cost of PV system with 5 days of Sodium Sulphur battery storage = $4,600 billion.

Solar Thermal

The system must be able to supply the power to meet demand at all times, even during long periods of overcast conditions. We must design for the worst conditions.

We’ll consider two worst case scenarios:

1. All power stations are under cloud at the same time for 3 days.

2. At all times between 9 am and 3 pm at least one power station, somewhere, has direct sunlight, but all other power stations are under cloud.

Assumptions:

The average capacity factor for all the power stations when under cloud for 3 days is 1.56 % (to be consistent with the PV analysis in “Solar Power Realities”; refer to Figure 7 and the table on page 10).

The capacity factor in midwinter, when not under cloud, is 15% (refer Figure 7 in “Solar Power Realities”).

Scenario 1 – all power stations under cloud

Energy storage required: 3 days x 450,000 MWh/d = 1,350,000 MWh

Hours of the day when energy is stored (9 am to 3 pm) = 6 hours

Average power to meet direct day-time demand = 25 GW

Average power required to store 450,000 MWh in 6 hours = 75 GW

Total power required for 6 hours (9 am to 3 pm) = 100 GW

Installed capacity required to provide 100 GW power at 1.56% capacity factor (say 6.24% capacity factor from 9 am to 3 pm) = 1,600 GW.

Total peak generating capacity required = 1,600 GW

If the average capacity factor was double, the installed capacity required would be half. So the result is highly sensitive to the average capacity factor.

Scenario 2 – at least one power station has direct sun at all times between 9 am and 3 pm

One power station provides virtually all the power. The other power stations are under cloud and have a capacity factor of just 1.56%.

Energy storage required for 1 day = 450,000 MWh

Hours of the day when energy is stored (9 am to 3 pm) = 6 hours

Average power to meet direct day-time demand = 25 GW

Average power required to store 450,000 MWh in 6 hours = 75 GW

Total power required = 100 GW.

The capacity factor in midwinter, when not under cloud, is 15% (refer Figure 7 in “Solar Power Realities”).

Installed capacity required to provide 100 GW power at 15% capacity factor (60% capacity factor from 9 am to 3 pm) = 167 GW.

But the clouds move, so all the power stations need this generating capacity.

To maximise the probability that at least one power station is in the sun we need many power stations spread over a large geographic area. If we have say 20 power stations spread across south east South Australia, Victoria, NSW and southern Queensland, we would need 3,300 GW – assuming only the power station in the sun is generating.

If we want redundancy for the power station in the sun, we’d need to double the 3,300 GW to 6,600 GW.

Of course the power stations under cloud will also contribute. Let’s say they are generating at 1.56% capacity factor. Without going through the calculations we can see the capacity required will be between the 1,600 GW calculated for Scenario 1 and the 3,300 GW calculated here. However, it is a relatively small reduction (CF 3% / 60% = 5% reduction), so I have ignored it in this simple analysis .

So, Scenario 2 requires 450,000 MWh storage and 3,300 GW generating capacity. It also requires a very much greater transmission capacity, but we’ll ignore that for now.

Costs of Solar Thermal with storage

NEEDS , 2008, “Final report on technical data, costs, and life cycle inventories of solar thermal power plants” Table 2.3, gives costs for the two most prospective solar thermal technologies. They selected the solar trough as the reference technology and did all the calculations for it. The cost for a solar trough system factored up to 18 hours storage and converted to Australian dollars is:

langsat1

This would be the cost if the sun was always shining brightly on all the solar power stations. This is about five times the cost of nuclear. However, that is not all. This system may have an economic life expectancy of perhaps 30 years. So it will need to be replaced at least once during the life of a nuclear plant. So the costs should be doubled to have a fair comparison with a nuclear plant.

In order to estimate the costs for Scenario 1 and Scenario 2 we need costs for power and for energy storage as separate items. The input data and the calculations are shown in the Appendix.

The costs for the two scenarios (see Appendix for the calculations) are:

langsat2

Summary of cost estimates for the options considered

langsat4

The conclusion stated in the “Solar Power Realities” paper is confirmed. The Capital cost of solar power would be 20 times more than nuclear power to provide the NEM demand. Solar PV is the least cost of the solar options. The much greater investment in solar PV than in solar thermal world wide corroborates this conclusion.

Some notes on cloud cover

A quick scan of the Bureau of Meteorology satellite images revealed the following:

This link provides satelite views. A loop through the midday images for each day of June, July and August 2009, shows that much of south east South Australia, Victoria, NSW and southern Queensland were cloud covered on June 1, 2, 21 and 25 to 28. July 3 to 6, 10, 11, 14. 16, 22 to 31 also had widespread cloud cover (26th was the worst), as did August 4, 9, 10, 21, 22.. This was not a a rigorous study.

Also see the BOM Solar Radiation Browse Service for March and April 2002 (the data on this site only goes up to 14 April 2002). Notice the low solar radiation levels for 25 to 30 March and 8 to 12 April 2002 over the area we are looking at. The loop animation can be viewed here.

Some comments on Future Costs?

How much cheaper can solar power be? NEEDS figure 3.7, p31 suggests that the cost of solar thermal may be halved by 2040.

How much cheaper can nuclear be? Hanford B, the first large reactor ever made, was built in 15 months, ran for 24 years, and its power was expanded by a factor of 9 during its life. If we could do that 65 years ago, for a first of a kind technology, what could we do now by building on experience to date if we wanted to put our mind to it. Is it unreasonable to believe that, 65 years later, we could build nuclear power plants, twenty times the power of the first reactor, in 12 months, for 25% of the cost of current generation nuclear power stations?

Appendix – Cost Calculations for Solar Thermal

The unit cost rates used in the analyses below were obtained from: NEEDS, 2008, “Final report on technical data, costs, and life cycle inventories of solar thermal power plants“, p31 and Figure 3.7.

langsat3

Note that, although this table includes calculations for the cost of a system with 3 and 5 days of continuous operation at full power, the technology does not exist, and current evidence is that it is impracticable. The figure is used in this comparison, but is highly optimistic.

———————————————–

Capital Cost of Transmission for Renewable Energy

Following is a ‘ball park’ calculation of the cost of a trunk transmission system to support wind and solar farms spread across the continent and generating all our electricity.

The idea of distributed renewable energy generators is that at least one region will be able to meet the total average demand (25 GW) at any time. Applying the principle that ‘the wind is always blowing somewhere’ and ‘the sun will always be shining somewhere in the day time’, there will be times when all the power would be supplied by just one region – let’s call it the ‘Somewhere Region’.

The scenario to be costed is as follows:

Wind power stations are located predominantly along the southern strip of Australia from Perth to Melbourne.

Solar thermal power stations, each with their own on-site energy storage, are distributed throughout our deserts, mostly in the east-west band across the middle of the continent.

All power (25GW) must be able to be provided by any region.

We’ll base the costs on building a trunk transmission system from Perth to Sydney, with five north-south transmission lines linking from the solar thermal regions at around latitude 23 degrees. The Perth to Sydney trunk line is 4,000 km and the five north-south lines average 1000 km each. Add 1,000 km to distribute to Adelaide, Melbourne, Brisbane. Total line length is 10,000km. All lines must carry 25GW.

Each of the double circuit 500kV lines from Eraring Power Station to Kemps Creek can transmit 3,250MW so let’s say we would need 8 parallel lines for 25GW plus one extra as emergency spare.

The cost of the double circuit 500kV lines is about $2M/km.

For nine lines the cost would be $18M/km.

So the total cost of a transmission system to transmit from the ‘Somewhere Region’ to the demand centres is 10,000km x $18M/km = $180 billion

The trunk transmission lines might represent half the cost of the complete transmission system enhancements needed to support the renewable generators.

Just the cost of the trunk transmission lines alone ($180 billion) is more than the cost of the whole nuclear option ($120 billion).

Eraring to Kemps Creek 500kV transmission line. Each of the double circuit 500kV lines from Eraring to Kemps Creek can carry 3250MW.  The 500kV lines are double circuit, 3 phase, quad Orange, i.e.2 circuits times 3 phases times 4 conductors per bundle, i.e. 24 wires per tower.  Orange is ACSR, Aluminium Conductor Steel Reinforced, with 54 strands of 3.25mm dia aluminium surrounding 7 strands of 3.25mm dia steel.  Roughly 1/3 of the cost of a line is in the wires, 1/3 in the steel towers and 1/3 in the easements required to run the line.

Eraring to Kemps Creek 500kV transmission line. Each of the double circuit 500kV lines from Eraring to Kemps Creek can carry 3250MW. The 500kV lines are double circuit, 3 phase, quad Orange, i.e.2 circuits times 3 phases times 4 conductors per bundle, i.e. 24 wires per tower. Orange is ACSR, Aluminium Conductor Steel Reinforced, with 54 strands of 3.25mm dia aluminium surrounding 7 strands of 3.25mm dia steel. Roughly 1/3 of the cost of a line is in the wires, 1/3 in the steel towers and 1/3 in the easements required to run the line.

Advertisements

Leave a Comment »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Create a free website or blog at WordPress.com.

%d bloggers like this: