The energy dynamics of energy production
When people talk about buying solar photovoltaic (PV) panels
they usually want to know how long it takes to repay the initial outlay with the subsequent
savings on electricity bills. This is called the financial pay-back time.
However if you are getting into PV because you want to help save the environment,
you should also be interested in the energy pay-back time - that is the amount of
time it takes the PV panels to repay the energy that was used in their manufacture.
This is important because it takes time before you can say your investment has made
an energy profit, and is therefore "helping to reduce the greenhouse effect".
Also, the price of energy can change, and what can make financial sense after government subsidies,
will not necessarily make 'energetic sense' in quite the same way.
To work out the energy pay-back time, someone needs to prepare an energy 'balance sheet', showing
all the energy inputs and outputs. This would include not only the electricity bill
at the PV factory, but also the embodied energy of all the materials used - purified silicon,
copper (for wiring), aluminium (for the frame), toughened glass (for the top plate),
lots of ultra-pure water and organic solvents,
and so on. On top of this, one also needs to know the energy spent in transporting the various materials
to the factory, from factory to retailer, and retailer to your house, and the energy
cost of building and equipping the PV factory.
The energy output depends on the nominal peak power of your PV panels (measured in Watts),
the lifetime of the panels (typically 25 years),
the location of your house, and the orientation of the panels on your roof. From these factors
it can be worked out how much energy will be captured by the panels over their lifetime.
For the PV panels available today, the energy output will be approximately
three times as much as the energy input. Opinions vary on the precise value of the
ratio Energy Returned over Energy Invested (ERoEI), depending on the scenario chosen
and the optimism of the person choosing the input values. (The manufacturers of PV panels
are, unsurprisingly, particularly optimistic about the thing they want to sell you.)
The numbers used in this article are only indicative, and are drawn from the work of
University of Sydney's ISA team [ 1 ] .
If the ERoEI of PV panels works out to be 3.0 , and the lifetime of the panel is 25 years,
then the energy pay-back time is 25/3 = 8.3 years, in other words it takes over 8 years
for the panels to pay back the energy used in their manufacture. Another way of expressing that
is to say that a PV panel can only pay back 12% each year of the energy needed to build it.
It is clear from this that a PV factory cannot be self-sustaining in energy until
it has been in operation for over 8 years, and until that point, it needs an energy subsidy
from another, presumably fossil-fueled, energy source or sources.
Modelling the PV factory's energy budget
How much fossil-fuelled energy does it take to establish a PV industry that is big enough
to have a substantial impact on the nation's energy mix ?
The dynamics of supplying energy to a growing PV industry does not seem to have been
studied before, and it produces some surprising, almost counter-intuitive results.
This study is based on a simple spreadsheet model, which you can
download from here .
However I am pitching this article at
an audience that will probably shy away from looking too deeply into the entrails of the model.
Consequently I am going to try and describe the model in plain English, and you only
need to understand the model if you want to try out your own scenarios.
Essentially what is happening in the model is that, using the example data above,
in the first year the PV factory will spend 8.3 units of energy on building a panel,
and then for each of the next 25 years, that panel pays 1 unit back. This gives us
a series of numbers : -8.3, +1, +1, +1, .... +1 which you can see in the spreadsheet
table highlighted in blue. (In practice, the PV factory will be building millions
of panels, but we will be scaling the production numbers up later.) The units used
are strictly "panel-years", that is, 1 panel operating for 1 year is counted as 1 unit.
In the second year the PV factory builds another panel, so the net energy profit for the second year
is -8.3 +1 = -7.3 units, that is a loss of 7.3 units. In the third year, the factory
makes another panel costing 8.3 units, and gets 2 units back, for a net loss of 6.3 units.
This process is continued for 50 years. In the tenth year, the energy cost of -8.3 units
is exceeded by the production of the 9 earlier panels, and the factory makes a net energy profit
for the first time. After 25 years, the panel made in the first year is assumed to die of old age and makes
no further contribution.
The calculations are summarised in a chart.
The 'no growth' scenario
In this scenario, the PV factory's production remains the same at 1 panel per year.
The blue line represents the energy profit for the current year, measured against
the blue scale on the right of the chart. You can see that it starts off at -8.3
and increases by one each year for 25 years. At that point, the first panel dies off,
and the new panel therefore only replaces the output of the old one, so from there
on the annual profit remains steady at +16.7 units.
The red line represents the cumulative energy profit/loss since the PV factory started,
measured against the red scale on the left of the chart. It starts off at -8.3 units
and dips lower and lower for 8.3 years, then rises for 8.3 years until it breaks even in 2024,
then it moves into positive territory, representing a real cumulative energy profit.
At its lowest point, the cumulative energy loss is 39 units. This means that if your factory
is making 1 million panels per year, it will need an energy subsidy that builds up
to 39 million panel-years by the ninth year, and isn't fully paid off until the seventeenth year.
This energy subsidy already takes the output of the panels themselves into account,
so it can only be supplied by some other energy source. This new demand for energy,
at a time when we are hoping to cut down on energy demand, represents an "energy barrier"
to the broadscale introduction of PV panels. If this energy barrier is ignored (as it is currently
being ignored) then everyone will be surprised to find that the big push for more
solar energy actually causes a big push for other kinds of energy in the short and medium term.
Scaling up the production level
The above example uses a PV factory with a production rate of 1 panel per year. How much
energy is that in familiar terms ?
Let us assume the panel is the largest one (and best value) currently available - rated at 175 Watts peak power,
and that it is located in Sydney (an average location for Australia) on a roof facing north
and tilted at an angle equal to Sydney's latitude, 34°, and taking average cloudiness into account. Under these circumstances, the panel
produces 1 KiloWatt.hour (KW.h) per day, so 1 'standard' panel-year is equal to 365 KW.h . Other locations
will produce different results - see [ 2 ].
Australia's electricity generation in 2006 was 257.8 TW.h [ 3 ] so
that is equivalent to 706 million panels. That was an increase over the 2005 figure of
8.2 TW.h (3.3%), so just the annual growth in electricity generation is equivalent to 22.5 million of our
standard panels.
As we have seen above, each panel requires 8.3 panel-years to build it, so a factory producing 22.5 million
panels will need an energy subsidy of 68 TW.h in the first year. This is equivalent to 26% of
our total electricity production.
I would suggest that it is impossible for the nation to divert 68 TW.h of energy
into PV factories merely so that they can build enough PV panels to meet the 3.3% growth in electricity
consumption. This is despite the fact that in the long term (more than 17 years ahead)
those panels will be making a handsome energy profit.
Production growth scenarios
Well, if it is not possible to start with producing 3.3% of Australia's electricity,
can we start smaller and grow the PV production capacity over time ?
The model allows us to enter a percentage growth per year factor, this is the chart from the scenario
with 5% growth :
As you can see, the annual profit now keeps growing, as when the first panel dies
of old age, it is replaced by more than one new panel, due to the growth in production
over the 25 years.
But note also that the cumulative energy break-even point has been pushed
out to 21 years, and the maximum deficit is 54 panel-years' worth of energy in the
12th year.
Because there are more panels being created in this scenario, you might think that
the results wouldn't have to be scaled up so much to meet the target, but what is
the target exactly ? The zero growth scenario doesn't make 100 panel-years profit
until 2032, but the 5% growth scenario has only made a 75 panel-year profit by 2032,
so if that is your target, then the growth model is worse. This is because
more of the energy produced by the PV panels is being ploughed back into production
in the growth scenario, and less is available as energy profit.
This might seem counter-intuitive, but the effect is real enough. And if the growth
is increased to 10% per year, we get this scenario :
Due to even more energy from the PV panels being ploughed back into new production,
the cumulative energy break-even point has now been pushed out to 2040, and I hope
you will agree that it is not wise to take on a project with such a
delayed energy profit, even if the energy profits from that point on are spectacular.
We are doing this to avoid fossil fuel emissions causing Climate Change, after all.
Since our PV panel can only repay 12% of the energy needed to build it each year,
any attempt to grow the PV production rate at more than that amount will result in a permanent
and increasing energy deficit :
So you see increasing production each year does not help solve the problem. The thing
that helps most is to stop producing panels altogether.
Improvinging the Lifetime factor
The model also allows the lifetime of the PV panel to be changed. This directly affects
the Energy Returned (ER) over the lifetime of the panel, and hence it alters the ERoEI.
However it does not affect the Energy Invested (EI), so the energy barrier, which
has to paid in the early stages of the project, remains the same.
Improving the ERoEI factor
The ISA model of PV production that gives an ERoEI of 3 (range 1.5 through 6.0) is based on the scenario
of a 100 MW solar farm, with associated electrical infrastructure, which will obviously
be pretty heavy-duty (energy-intensive) equipment. Other scenarios will give different results for ERoEI. Even so,
an ERoEI of 6 and Growth of 5% still has a 10 year wait before a Cumulative Energy Profit
is achieved.
Application to other energy sources
With PV solar, all the Energy Invested over the lifetime of the panel is invested
up front, before any Energy Returned is seen. However other energy sources, particularly
those needing fuel or on-going maintenance or expensive decomissioning, some of the EI
is spent over the lifetime, and only a proportion spent up front.
In my next article I shall be introducing Energy Invested Up Front (EIUF) and the ratio EIUF/EI,
which is 100% for solar PV. With suitable modifications to the model, and drawing on the
ISA Team's modelling data, we can look at other energy sources in the same way.
Conclusion
We have been living in an era of expanding energy availability, but Peak Oil and the
constraints of Global Warming mean we are entering a new era of energy scarcity.
In the past, you could always get the energy you wanted by simply paying for it.
From here on, we are going to have to be very careful about how we allocate energy,
because not only is it going to be very expensive, it will mean that someone else
will have to do without. For the first time, ERoEI is going to be critically important
to what we choose to do. If this factor is ignored, we will end up spending our fossil energy
on making solar energy, which only makes Global Warming worse in the short to medium term.
Dave Kimble
References
1. "Life Cycle Energy Balance and Greenhouse Gas Emissions of Nuclear Energy in Australia"
Integrated Sustainability Analysis team, University of Sydney, November 2006
http://www.pmc.gov.au/umpner/docs/commissioned/ISA_report.pdf [2.75 MB]
2. Australian insolation data
Excel speadsheet based on UML's Solbase database
3. BP's Statistical Review of World Energy (2007)