Highlights
- On a global level, the potential for renewable energy is more than sufficient.
- Problems emerge on a regional level, however, especially in developing Asia and Africa.
- Renewable energy technology forcing in these regions can have serious socio-economic consequences.
Introduction
We often see images like the one below which imply that the potential of renewable energy is essentially limitless. Thus, if we only had the will, we could easily power the world with clean and everlasting renewable energy.
Solar PV is generally viewed as the most limitless of all the renewable energy options. The little squares on the map below shows just how easy it is to power the world with solar.
The reality is, however, that realistic renewable energy potential is some orders of magnitude lower than these simplistic illustrations.
Firstly, areas covered by urban developments, forests, protected zones, ice, dunes or rock need to be excluded. In addition, areas with excessive slope or elevation are also not eligible build sites. After these eliminations, only areas with a sufficiently strong solar irradiation and wind speeds can be considered.
From the remaining land area, only a small fraction can be used before serious social resistance or natural habitat interference is encountered. For example, only about 1-2% of available land area is covered by onshore wind in European countries like Denmark, Germany and the Netherlands, but these issues are already becoming significant.
All of these factors have recently been quantified in a very interesting study published in the Elsevier journal “Global Environmental Change”. Findings from this study are further discussed below.
Globally – more than enough
Even after all of these realistic assumptions, the total global wind and solar resource still easily meets projected demand by the year 2070 even under the most pessimistic assumptions (the dark bands in the graphs).
It is clear that PV, CSP and offshore wind hold the greatest potential. Onshore wind has a much smaller potential, however, especially under low (3%) and medium (6%) land availability assumptions. PV on buildings also has quite a large potential in the year 2070 due to assumptions of large urban buildouts and large gains in solar panel efficiency (35% in 2070).
The projected electricity demand by 2070 is set within the range of 24-40 GJ/person/year. For perspective, the average American currently consumes about 44 GJ of electricity per year and electricity accounts for only about 20% of final energy consumption.
Regionally – problems arise
Unlike hydrocarbon fuels, electricity is not easily tradeable between different world regions. It is therefore very important to assess renewable energy resource availability on a regional basis. The following highly informative graphic tells the story:
It is clear that only North America, developing Europe and Australia have access to a well-balanced mix of renewable energy resources with more than enough potential. A well-balanced mix of resources is important to minimize the effects of intermittency in order to allow for higher renewable energy market shares. For example, the positive effect of mixing wind and solar in terms of preserving more value with increasing market share is shown below (the y-axis illustrates the value of generated electricity where 1 is the average market value):
When deploying only wind or only solar PV, the solar PV option is especially challenging. Because solar’s variability is very pronounced and highly correlated within a reasonable distance, its value falls rapidly with increasing market share. This is illustrated below:
Offshore wind and rooftop solar are about twice as expensive as onshore wind and utility-scale PV for obvious reasons. In addition, the development of CSP has been slower than anticipated. It should be mentioned, however, that the low-cost inclusion of thermal energy storage in CSP significantly increases its value.
Given these considerations, most regions around the world will have a very tough time achieving high market shares of renewable energy. The two most populous regions in 2070: Sub-Saharan Africa and South Asia will have to rely heavily on solar power. If solar thermal technology can be greatly improved, this will help Sub-Saharan Africa, but South Asia will have to rely almost completely on solar PV – mostly the expensive distributed kind. North Africa and the Middle East face similar challenges.
The highly populous East Asian, South-East Asian and South American regions can achieve greater balance if they heavily rely on expensive offshore wind. South-East Asia will be especially dependent on offshore wind together with EU Europe.
It should be noted that further refinement of the data to a country or state level will further accentuate these challenges. The situation outlined above assumes lots of long distance electricity lines and excellent performance by politicians to establish cross-border regional electricity markets.
Special challenges for the developing world
The challenges outlined above are further augmented in the developing world, especially Asia and Africa which may well be home to 80% of the world population by the end of this century:
These regions and their enormous populations still have a lot of industrialization to do. Industrialization is critical to give these people a reasonable quality of life, to shield them against the effects of climate change, and to naturally curb population growth. Unfortunately, industrialization is also an incredibly expensive and resource intensive undertaking. Insisting on driving industrialization primarily through renewable energy will therefore come at a tremendous cost in terms of quality of life, especially given the challenges outlined in the previous section.
As a simple example, I estimated the effects of renewable energy technology forcing on economic growth in India as an example at the bottom of this article. The example showed that deployment of only solar and wind to grow Indian electricity production to support economic growth would cut the Indian growth rate in half. After 20 years of this practice, the Indian economy would literally be only half the size it could otherwise have been. This situation will be further worsened given the fact that South Asia will have to rely heavily on expensive distributed solar PV which will rapidly lose value as market share increases. Such a development strategy is simply not going to happen unless rich nations finance the necessary subsidies. And that is not going to happen any time soon.
Final word
This article was definitely not written to write off renewable energy. As often stated before, I wholeheartedly support moderate wind and solar deployment in regions where they make sense. For example, the US is one country where renewable energy makes a lot of sense due to its vast available land areas, high quality wind and solar resources, and affluent population.
Wind and solar technology forcing in regions with much lower potential and much poorer populations is a completely different story though. I fear that this strategy will be highly inefficient at best and disastrous at worst.
“Unlike hydrocarbon fuels, electricity is not easily tradeable between different world regions”
Skip “unike” and “not” in this sentece when considering today’s and tomorows HVDC-Technology.
Already today a power line with a length of 3284km and a capacity of 12GW on a single system is under construction in Chna.
This distance is enough to connect Canada with Ireland, or the end of Alaska with Sapporro in Japan. And it is not the end of the development.
…wouldn’t it be prudent to wait until one of these gigantic infrastructure projects is completed and interconnected before touting a technology which has to be tested in the field? Or better yet, wait until this technology is deployed outside of China? That’s not to mention the political and social challenges around siting large scale transmission projects, a component of energy systems which is very charged in the U.S.
There are a lot of HVDC lines in china in use which are just a bit shorter in length and in the power transfered.
So it it is nothing really new, just incremenatl improvement. Just many do not notice it since it is not happening in front of the own door. The transmission line itself does not look significant different from existing 800kV AC transmission lines. Just instead of 2 AC systems ther could be 3 HVDC Systems, and the power transmitted per wire can be significant higher.
There are also other comodities like food which are not durable, not easily storable but traded nevertheless. I would also not describe natural gas as “easily storable”. It is storable with big efforts, similar to storages in hydropower stations, just mostly underground.
There was only one offshore project which had significant delays, because there was a prototype of HVDC-equipment in use designed by a company with no HVDC-experience. The delay is negible comparing to Vogtle or Sumner.
And the North-South corridors – there are already several with >20GW capacity, the oldest existing nearly a century. The new ones are no technical challenge, it is just a question of red tape, which is a problem for any big project, so also for conventional power stations in germany today. They will be operational on time, as usual, but about 2-3 years behind original shedule. Project costs are mainly red tape costs for power lines here.
Easily tradable commodities are durable, portable and easily storable. Electricity is none of these.
Connecting one relatively small country can be a challenge.
Your native Germany has had big problems connecting offshore wind farms (distance 30-60 miles) to the grid. And the Stromautobahn connecting north and south Germany will be ready when? Just 500 miles of powerlines. And the projected cost?
Future large scale inter-regional renewable energy exports will have to rely heavily on solar PV. For this reason, we can expect a requirement of several TW of HVDC capacity over distances approaching 10000 km with a utilization rate of only about 15% (only the solar power peaks will be exported). In practice, this could be useful in some cases such as the Middle East and North Africa sending solar power to East and South Asia to cover their afternoon/evening peak in electricity demand.
However, this sounds expensive (massive capacity and distance combined with low utilization), complex (have to cross multiple countries with very different political agendas as well as very complex terrain) and restrictive in terms of market (exporting East is best). I think it would be much easier and more profitable (but still quite expensive) for the Middle East and North Africa to simply convert their excess solar power to synfuels and sell that on the global market.
There are some calculation errors. Flow in interconnectors (grids) runs in two directions, this doubles the utilisation of the interconnector. Also a day does not just last 24×0,15=3,6 hours
Since losses in a interconnector are roughly according to the square of the power transfered, the transfer starts with a lower utilisation, and according to this with low losses at a close to zero price difference. It will be ramping up with higher price differences. Interconnectors also balace out wind power, and differences of demand.
Converting to synfuels would not cost significant less as far as equipment is concerned, but produce significant higher losses, and provide a less valuable product. Synfuel production would also profit from large and strong grids, being able to access excess electricity from multiple areas thus allowing a far better utilisation of the equipment.
To cross several countries is not so much a problem as pratice tells, if each country gets access points to the lines, and thus profits from the stable energy supply the lines provide.
Be aware that e.g. marocco, algeria, tunesia, and as it seems also lybia, egypt sudan and Ethiopia are running synchrouns to the european grid, while kenia uganda and the other states of the east african power pool run synchronus with south africa (south african power pool). A 2 GW HVDC connection to connect both non synchrounus areas between Kenia and Ethiopia is under construction. (A second connection is under development from South Africa to grand Inga in Kongo, there connecting towards the west african power pool, where constructions are under way to connect to Marocco, as it seems this connection will be synchronus with the european grid, thus synchronising the west african power pool with the european grid) So crossing borders seems to be a smaller problem, at least in this area.
And it is no problem when running cables across a ocean. (like e.g. the eurasia connector, connecting greece, crete, cyprus to israel, which is under develpment, or the greece-egypt interconnector, the sicilia-tunesia interconnector, the sardinia-tunesia interconnector, the portugal-maroccco interconnector etc.
I’m talking here about exports from a region with excess renewable energy potential to a region with too little renewable energy potential. The energy flow will therefore be mostly one directional.
Also, only excess energy will be exported (or converted to synfuels). This will be mostly excessive solar power in the middle of the day. If the export line is to be used at more than 15% capacity, some additional energy storage will be required.
This is also the argument for synfuels. Since the large amount of excess solar power will have a very low value, the low conversion efficiency is not very important. The most important cost will be the capital which will also be high due to the low utilization rate.
First the “excess” and “too little” changes with the time of day at both ends of the cable. The 10-20% differences in average yearly solar power production are much too small compared to the changes during the day to make such a cable a one way connection. There are about 8 hours difference between marokko and China, so for about 8 hours (Night and morning in Marokko) there will be a flow of power from the east end to the west, with the source of power mowing west, and then some time of undecided flow, and then a flow fromw est to east, during evening and night in china. Overlying to this the varying patterns of wind power generation all along the line, and of hydropower from one place supplying power from another place, or storing power from another place.
I acknowledge that inter-regional HVDC lines can have some additional utility in balancing intermittent renewables. However, the numbers of deploying a 10000 km HVDC line for arbitrage don’t work. If we assume optimistic costs of $200/MW/km (bit more than half of current US costs on flat and free land) and ignore substation costs, a 1 GW 10000 km line will cost $2 billion. Accounting for losses, this works out to about $3000/kW capacity. Even assuming no O&M costs, zero value of the losses and no interest during construction, the levelized cost over 40 years at a 7% discount rate and 50% capacity factor is $51/MWh. The market will have to be quite crazy to present wholesale arbitrage opportunities of this magnitude anywhere near 50% of the time. And more realistic cost assumptions may well double the levelized cost assumed in this example.
If we shorten the line to 1000 km, the numbers improve substantially. HVDC for intra-regional balancing (primarily of wind) can therefore make sense. But the distances required for inter-regional balancing of primarily solar PV are simply too great.
Renewable energy has room to grow on the margins of our energy systems, harvesting the low hanging fruit. But renewables are as far away as ever in offering solutions for the most challenging portions of our energy systems, namely heavy duty transportation and high temperature industry.
You can install all the solar panels you want, but you can’t run a solar panel factory with them much less an Air Force.
I’ve heard the number of 100 GJ/year/person final energy consumption to ensure a global society of reasonable material standards. This analysis only assumes 24-40 GJ/year/person of electricity, so hydrocarbon fuels should still play an important role.
Of course, the ideal of 100% wind/solar/hydro that some people have in mind will require much more than 100 GJ/year/person in electricity output because a large chunk of electricity will have to be converted into some form of synfuel at a substatial thermodynamic loss.
Increasing the projected electricity demand by a factor of 4 from the numbers in this study really highlights the restrictions of renewable energy in developing Asia and Africa.
I have paid attention to this very fact. Electricity is supposed to replace fossil fuels in transportation, heating and industrial processes.
Yet for example German Energiewende goal is to consume less electricity in the future. Using 2008 as the base year (100%), 2050 goal is 75%, of which 80% is of renewable origin.
Of course Energiewende 2050 goal still has a lot of room for fossil fuels. 40% of the gross energy consumption would still come from fossil fuels, 60% from renewables.
100GJ per person is not much. In my native Finland the consumption in 2015 was 252 GJ per person in 2015. Of which 45% came from fossil fuels. So we are already getting well over 100GJ per person annually from renewables and nuclear.
Yes, a comfortable life will probably require more than 100 GJ/year, but reaching this level in a given country will at least remove the worst effects of poverty.
In a future highly electrified energy system, a 100 GJ/year of final energy consumption should also go a lot further than it does today. Of course, if we want all energy to come from electricity, efficiency will drop again due to the need to make synfuels.
Good points. But there currently appears to be little hope for a recovery of the CSP industry. PV has greatly overtaken CSP, and will likely grow until daytime electricity is cheap, which will block further investment in all solar.
But given that thermal energy storage is so much cheaper and cleaner than batteries, it is worth noting that several kinds of Gen IV nuclear power operate at a suitable temperature for thermal energy storage (e.g. the sodium cooled reactors which are being prototyped in Russia, China, and India, as well as the Chinese PBMR and salt cooled reactors). Plus, unlike CSP, nukes have the added advantage of producing most of their output when PV is not producing.
Well, when transmission is combined open markets, and the uneven global distribution of renewable resources, it seems likely that the low cost producers will force out other providers. How can cloudy Germany hope to compete against Middle Eastern solar exports? And given that electricity is enormously less storable than fossil fuel, that will provide even more reason for developed countries to maintain a military presence in the Middle East to insure the uninterrupted flow of energy.
Interesting point about nuclear thermal storage.
Well, I guess the best argument for Germany to use German PV instead of Middle East PV is that getting the Middle East PV to Germany will be very complex and expensive. As a long-term alternative, the Middle East might become a big exporter of PV-derived synfuel which will be much easier to export.
No, because the differences in prices is not wotrth military presence. Hydropower in Island is even cheaper than solar power in middle east today, and there is no need for a military presence there.
The only point is, to have connections in many directions, and so to trade with many partners, and not to rely on a single source.
Spain, southern italy and greece are sunny enough to compete with middle east due to lower transaction costs. It’s like the oil and gas industry in the US which kept pumping although the saudis could pump oil for much less.
If all uranium producing countries place a embargo on your state, nuclear also gets in trouble. Same for coal, oil and gas. There are much less countries from which you can buy significant amounts of gas than there are countries from with which you can trade renewable electric power in both directions.