Q&A: First Solar's Jack Curtis - Part 2

In the second part of our interview with First Solar's Jack Curtis we delve into the benefits of large-scale solar and why it is worthy of government support.

In this second part of our interview with Jack Curtis of First Solar, the company that will provide the technology for AGL’s 159MW Solar Flagship projects, we focus on the issue of why it’s worthwhile doing large-scale, wholesale market solar PV projects, when residential and commercial consumers pay so much more for their electricity than the wholesale market.

Curtis raises some excellent points about why large-scale solar is worthwhile supporting and offers good value relative to support for smaller residential systems. While I disagree with some of the points he makes as they might relate to commercial rooftop systems capturing the value of network costs, he makes a persuasive argument for the importance of supporting large-scale solar. 

The key thing that came out of this interview that needs highlighting is that I made an incorrect comparison between the cost of residential PV systems and the AGL-First Solar project in the Climate Spectator article on Tuesday.

The figure I cited for cost per watt of residential systems is on a direct current generating capacity basis whereas the 159 megawatt figure for the AGL First-Solar Project is alternating current. On a direct current basis the AGL-First Solar projects represents 200 megawatts, which equates to a cost per watt of $2.25, not the $2.80 I originally calculated. This is noticeably below what the market is offering for smaller scale solar PV systems. In addition, this large project is likely to generate noticeably more electricity than most residential systems, due to better capture of the solar energy available from the site and higher efficiency of the overall system engineering.

Nonetheless, this does not change the fact that the Solar Flagships process has been an exercise in how to not go about driving uptake of solar in this country. Firstly, it will still take five to six years before we’ll see solar electricity delivered to the grid, which is simply unacceptable. And even taking into account the lower installed cost per direct current watt and higher energy yield, the government subsidy provided per megawatt hour for the Solar Flagships projects (assuming a long-run Large-Scale Renewable Energy Certificate price of $55) is not a meaningful improvement over what is provided for residential systems.

This doesn’t mean government policy should therefore focus on residential systems and not large-scale. In fact what it means is that government policy should stop picking and choosing what size of system it will support, and instead be structured around a subsidy per unit of solar electricity irrespective of its source. This should hopefully lead to solar systems being installed where the value of their electricity is most valuable to the customer and ultimately enable us to get the biggest bang for our buck.


TE: One of the standout issues with this particular project is that $2.80 per watt is not necessarily all that much better than what some of the local industry are delivering for much smaller scale projects, particularly at a residential and a commercial sector level.  Which leads to a question about why do these big projects?

JC: So, I think there are a couple of answers. The first one is that that comparison to residential system price is comparing a residential system at $2.80 per watt on a Direct Current basis against an Alternating Current figure. So cost per watt for our project is probably about 30 per cent less than that. 

I think the other point I’d make is that residential has been very well rolled out in this country over the past three or four years, so they have managed to achieve a lot of those localised cost reductions through installation methodology and design optimisation which we’re really just at the beginning of as it relates to utility scale solar. So, I think there’s a lot more cost reductions that can be captured for large scale just as the residential model has done a great job of capturing cost reductions over the past four or five years. 

The other main thing to bear in mind is at the end of the day we’re selling kilowatt hours, not kilowatts, and so the yield that you get out of a ten kilowatt distributed or rooftop system is generally lower than a thousand ten kilowatt utility scale systems all on a vacant field. 

So if you look at where the actual cost of electricity coming out of a PV system today is and where the trajectory is going over the next ten years, I think the APVA’s latest numbers had residential grid parity at around twenty-seven cents a kilowatt hour, you know, trending down to a bit under twenty cents a kilowatt hour over the next ten years which I think is a very sensible and attractive goal.

But you compare that with this project for example which is, as I said, the first project and inherently more inefficient than subsequent projects will be, that’s delivering electricity unsubsidised at about a hundred and eighty dollars a megawatt hour already. And so, then the next thing you have to overlay is well what are you competing against?

So residential systems on a cents per kilowatt hour number even though they generate electricity at a higher cost, they’re competing with a higher benchmark which are retail rates compared with utility scale which needs to displace wholesale rates.  And so, when you do that comparison, you’re essentially seeing, you know, retail electricity bills anywhere from, you know, high twenties down to, you know, low teens if you’re a large commercial industrial user.

But then you have to look at well how much of that are you actually offsetting.  So, in anyone’s retail bill there are various components of what comprises that price and some of that is the variable cost of the electricity, but some of it also includes a meaningful component associated with fixed transmission and distribution charges.  And when you put a rooftop system on a roof, you’re not offsetting those fixed transmission distribution charges.  You’re only offsetting the variable electricity cost whereas when you install a grid connected utility scale system, you’re offsetting the entire, you know, wholesale benchmark that you’re trying to compete against. 

So, when you’re installing at $100-$140 a megawatt  hour, that’s the goal you’re trying to get to, if you can install a system at that price, you’re offsetting the entire price.  If you install a residential or commercial rooftop system at 20 cents a kilowatt hour, you’re not offsetting the entire 20 cents a kilowatt rate. 

So, I guess the three main points are:

1) It’s not quite an apples to apples system price comparison;

2) The real focus is on cost per kilowatt hour and a yield and financing and development and transaction cost benefit at utility scale solar that obviously don’t get realised at a smaller scale; and

3) You need to compare what are you generating against what are you offsetting and how much of that are you offsetting? 

But all that’s not to say that we don’t support residential. 

You know, we think residential and distributed has a very big role to play in the energy mix.  We think it has a very strong chance of being able to get to a retail delivered parity price level and just because we don’t play in that space it’s not because we don’t think it has a lot of potential.  

We just think that utility scale, you know, has a lot of potential for different reasons and, you know, is without doubt the cheapest way of generating solar electricity.  You just need to make sure it’s offsetting the benchmark that you need it to versus the benchmark the retail or distributed solar PV is trying to offset.

TE: Just getting to that sort of apples for apples comparison, one of the useful things is to look at capacity factor from the system and you might do that on a basis of say a DC watt installed or an AC watt installed.  What sort of capacity factor are you expecting for the projects that are being installed in Nyngan and Broken Hill?

JC: Yeah.  It’s about 27 per cent. 

[Editor’s note: this was on the basis of the 159MW alternating current figure. The capacity factor is just under 21 per cent if you use the direct current rated capacity of these projects and would compare to about 14-15 per cent for solar PV residential systems installed in Australian capital cities]

TE: Are they on trackers or are they just fixed?

JC: It’s just fixed tilt.

TE: Does that take into account the losses in converting to AC?

JC: Yes. That’s an AC capacity factor.

TE: That’s delivered onto the transmission line before line losses?

JC: That’s right. And what that kind of supports, Tristan, is that point I was making before, you know, where cost reduction has been achieved in other markets at around the optimisation of these plants.  So, when you’re dealing with a rooftop, there’s not a lot you can do with orientation, shading, you know, the way that you angle the panels.  You know, you’re very constricted by the space you have to work with and when you’re working, you know, in a big field, there’s a lot you can do around that optimisation that increases that yield output.

TE: Okay, very interesting.  Jack, is there anything else that you wanted to cover off at all?

JC: No.  I think there are actually some very topical questions you’re asking, Tristan, hitting all the right themes.  I think that, you know, one of the main things we’re trying to focus on is getting the industry to commit to a point in the not too distant future where we can deliver these projects without subsidies and if we can’t, then you know frankly I think we deserve to be relegated to the niche technology heap again.  And so, what we’re trying to do is communicate a) where is the real price of this technology today, b) where do we think it needs to get to, and c) how do we get there?  And we think there are some fairly credible things that can be done without too much of a burden on the taxpayer, on the government that gets us to that point where, you know, we don’t need to be doing these projects with government support.

Related Articles