Category Archives: Climate change policy

Europe’s phase out of coal

Europe is progressing with phasing out hard coal and lignite in power generation, but needs to move further faster, especially in Germany and Poland

Reducing coal use in power generation and replacing it with renewables (and in the short run with natural gas) remains one of the best ways of reducing emissions simply, cheaply and quickly at large scale.  Indeed, it is essential to meet the targets of the Paris Agreement that the world’s limited remaining cumulative emissions budget is not squandered on burning coal and lignite in power generation.

Europe is now making progress in phasing out coal.  The UK experience has already illustrated what can be done with incentives from carbon pricing to reduce coal generation.  Emissions from coal have reduced by more than 80% in the last few years, even though coal plant remains on the system[i].  However, many countries, including the UK, are now going further and committing to end coal use in power generation completely in the next few years.  The map below shows these commitments as they now stand.  Most countries in western Europe now have commitments in place. (Spain is an exception.  The government is expecting coal plant to be phased out by 2030, but currently does not mandate this.)

Map: Current coal phase-out commitments in Europe[ii]

Source: Adapted from material by Sandbag (see endnotes).

In some countries there is little or no coal generation anyway.  In other countries plants are old and coming to the end of their life on commercial grounds, or are unable to comply with limits on other pollutants.  In each case phase-out is expected to go smoothly.

However, the largest emitters are mainly in Germany and Poland and here progress is more limited.  Germany has now committed to coal phase-out.  But full phase-out might be as late as 2038.  Taking another 20 years or so to phase out such a major source of emissions is simply too long.  And Poland currently looks unlikely to make any commitment to complete phase out.

This means the Europe is still doing less than it could and should be doing to reduce emissions from coal and lignite.  As a result, EU emissions are too high, and the EU loses moral authority when urging other nations, especially in Asia and the USA, to reduce their emissions further, including by cutting coal use.

Several things are needed to improve this situation, including the following.

  • Further strengthening the carbon price under the EUETS by reducing the cap. I looked at the problem of continuing surpluses of allowances in another recent post, and accelerated coal closure would make the surplus even greater.  Although the rise in the EUA price in the last 18 months or so is welcome, further strengthening of the EUETS is necessary to reduce the risk of future price falls, and preferably to keep prices on a rising track so they more effectively signal the need for decarbonisation.
  • Continuing tightening of regulations on other pollutants, which can improve public health, while increasing polluters’ costs and therefore adding to commercial pressure to close plant.
  • Strengthening existing phase out commitments, including be specifying an earlier completion date in Germany.
  • Further enabling renewables, for example by continuing to improve grid integration, so that it is clear that continuing coal generation is unnecessary.

As I noted in my last post, making deep emissions cuts to avoid overshooting the world’s limited remaining carbon budget will require many difficulties to be overcome.  There is no excuse for failing to make the relatively cheap and easy reductions now.   Reducing hard coal and lignite use in power generation in Europe (and elsewhere) continues to require further attention.

Adam Whitmore – 18th June 2019

[i] See https://onclimatechangepolicydotorg.wordpress.com/2018/01/17/emissions-reductions-due-to-carbon-pricing-can-be-big-quick-and-cheap/

With and updated chart at:

https://onclimatechangepolicydotorg.wordpress.com/carbon-pricing/price-floors-and-ceilings/

[ii] Map adapted from Sandbag:

https://sandbag.org.uk/wp-content/uploads/2018/11/Last-Gasp-2018-slim-version.pdf

and data in:

https://beyond-coal.eu/wp-content/uploads/2018/11/Overview-of-national-coal-phase-out-announcements-Europe-Beyond-Coal-November-2018.pdf

and https://www.eia.gov/todayinenergy/detail.php?id=39652

How well is the UK on track for zero emissions by 2050?

By 2020 the UK will have very nearly halved its emissions over 30 years.  Reducing emissions by the same amount over the next 30 years will get the UK very close to zero.  However this will be very much more difficult.

A robust net zero target has been recommended for the UK …

A recent report by the UK’s Committee on Climate Change (CCC), the Government’s official advisory body, recommends that the UK adopts a legally binding target of net zero emissions of greenhouse gases by 2050[i], that is remaining emissions must be balanced by removal from the atmosphere.  If the Government agrees, this will be implemented by amending the reduction mandated by the Climate Change Act, from an 80% reduction from 1990 to a 100% reduction.

The target has several features that make it particularly ambitious.  It:

  • sets a target of net zero emissions covering all greenhouse gases;
  • includes international aviation and shipping;
  • allows no use of international offsets; and
  • is legally binding.

This is intended to end the UK’s contribution global warming.  It has no precedents elsewhere, although in France a bill with comparable provisions is under consideration[ii].

Progress to date has been good …

The UK has made good progress so far in reducing emissions since 1990.  Emissions in 2018 were around 45% below 1990 levels, having reduced at an average rate of about 12.5 million tonnes p.a. over the period.  On current trends, over the thirty years from 1990 to 2020 emissions will be reduced to about 420 million tonnes p.a., 47% below their 1990 levels.  Emissions will thus have nearly halved over the 30 years 1990 to 2020, half the period from 1990 to the target date of 2050.

Chart 1 shows how the UK’s progress compares with a linear track to the current target of an 80% reduction, to a 95% reduction and to a 100% reduction.  (For simplicity I’m ignoring international aviation and shipping).  The UK is currently on a linear track towards a 95% reduction by 2050.

Chart 1: Actual UK emissions compared with straight line progress towards different 2050 targets

 

Source: My analysis based on data from the Committee on Climate Change and UK Government.  Data for 2018 is provisional[iii]

The largest contributor to the total reduction so far has been the power sector.  Analysis by Carbon Brief[iv] showed that the fall in power sector emissions has been due to a combination deploying renewables, which made up about of third of generation in 2018, reducing coal use by switching to natural gas, and limiting electricity demand growth.

Industrial emissions have also fallen significantly.  However some of this likely represents heavy industry now being concentrated elsewhere in the world, so likely does not represent a fall in global emissions.  Emissions from waste have also fallen, due to better management.

Reducing emissions will be relatively easy in some sectors …

There are also reasons for optimism about continuing emissions reductions.  Many technologies are now there at scale and at competitive prices, which they were not in previous decades.  For example, falling renewables costs and better grid management, including cheaper storage, will help further decarbonisation of the power sector.  Electrification of surface transport now appears not only feasible, but likely to be strongly driven (at least for cars and vans) by economic factors alone as the cost of batteries continues to fall.

But huge challenges remain …

Nevertheless important difficulties remain for complete decarbonisation.

CCS is identified by the report as an essential technology.  However, as I have noted previously, it has made very little progress in recent years in the UK or elsewhere[v].  CCS is especially important for decarbonising industry.  This includes a major role for low carbon hydrogen, which is assumed to be produced from natural gas using CCS – although another possibility is that it comes from electrolysis using very cheap renewables power, e.g. at times of surplus.  CCS also looks to be necessary because of its use with bioenergy (BECCS), to give some negative emissions, though the lifecycle emissions from this will require careful attention

Decarbonising building heating, especially in the residential sector, continues to be a challenge.  The report envisages a mix of heat pumps and hydrogen, perhaps in the form of hybrid designs, with heat pumps providing the baseload being topped-up up by burning of hydrogen in winter.  I have previously written about the difficulties of widespread use of heat pumps[vi], and low carbon hydrogen from natural gas with CCS is also capital intensive to produce and therefore expensive to run for the winter only.  The scale of any programme and consumer acceptance remain major challenges, and the difficulties encountered by the UK’s smart meter installation programme – by comparison a very simple change – are not an encouraging precedent.

Emissions from agriculture are difficult to eliminate completely, and no technologies are likely to be available by 2050 that enable aviation emissions to be completely eliminated.  This will require some negative emissions to balance remaining emissions from these sectors.

Policy needs to be greatly strengthened …

Crucially several of the necessary transformations are very large scale, and need long lead times, and investment over decades.  There is an urgent need to make progress on these, and policy needs to recognise this.  This includes plans for significant absorption from reforestation, as trees need to be planted early enough that they can grow to be absorbing substantial amounts by 2050.

The UK’s progress on emissions reduction so far has been good, having made greater reductions than any other major economy[vii].  And technological advances in some areas are likely to enable substantial further progress.  However much more is needed.  In particular policy needs to look now at some of the difficult areas where substantial long-term investment will be needed

Adam Whitmore – 22nd May 2019

 

 

[i] https://www.theccc.org.uk/2019/05/02/phase-out-greenhouse-gas-emissions-by-2050-to-end-uk-contribution-to-global-warming/

 

[ii] The CCC report notes that Norway, Sweden and Denmark have net zero targets, but they allow use of international offsets (up to 15% in the case of Sweden).  France has published a target similar to the UK’s in a bill.  The European Commission has proposed something similar for the EU as a whole, but this is a long way from being adopted. California has non-legally binding targets to achieve net zero by 2045.  Two smaller jurisdictions (Costa Rica, Bhutan) have established net zero targets but these are expected to be achieved mainly by land use changes.  New Zealand has a draft bill to establish a target, but eliminating all GHGs will be difficult because of the role of agriculture in the New Zealand economy.

 

[iii] https://www.gov.uk/government/statistics/provisional-uk-greenhouse-gas-emissions-national-statistics-2018  The change from 2017 to 2018 is applied to the data series from 1990 produced by the CCC (the two data series differ very slightly in their absolute levels).

 

[iv] https://www.carbonbrief.org/analysis-uk-electricity-generation-2018-falls-to-lowest-since-1994

 

[v] https://onclimatechangepolicydotorg.wordpress.com/2018/04/25/a-limited-but-important-medium-term-future-for-ccs/

 

[vi] https://onclimatechangepolicydotorg.wordpress.com/2015/05/18/reducing-the-costs-of-decarbonising-winter-heating-needs-to-be-a-priority/

 

[vii] https://onclimatechangepolicydotorg.wordpress.com/2017/05/09/uk-emissions-reductions-offer-lessons-for-others/

 

The EUETS has not been fully fixed

The reforms introduced to the EUETS for Phase 4 improve its functioning, but without further reform a chronic surplus looks likely and the risk of low prices remains.

The changes to the EUETS that were agreed in late 2017 make significant improvements to its design.  The temporary doubling of the intake rate for the MSR will reduce the surplus in the market more quickly.  And the provision to cancel allowances from the MSR when it exceeds a defined size will avoid the number of allowances in the MSR growing indefinitely.  The price of EUA’s has risen, although they remain below the levels needed to stimulate many efficient emissions reductions.  These changes have led some to conclude that the problems with the EUETS have been resolved.

However, major risks remain.  The cap for Phase 4 (which runs through the 2020s) was set on the basis of an overall reduction in emissions from 1990 levels of 40% by 2030[i].  In practice, emissions now look likely to reach around 50% below 1990 levels by 2030, and possibly to go lower than this if additional policies are put in place.  This looks likely to result in emissions remaining well below the cap throughout Phase 4.

This is illustrated in Chart 1 below, which shows three scenarios included in a recent report by climate NGO Sandbag[ii] (to which I contributed).  The correspond to overall reductions from 1990 levels of 50%-58% by 2030, rather than the 40% reduction on which the cap was set.

Many of the additional emissions reductions are from the sectors covered by the EUETS.  In particular increased renewables and decreased coal and lignite burn in power generation are the largest contributors to reduced emissions.  Consequently, in each scenario emissions remain well below the cap throughout the 2020s.

Even the European Commission’s own modelling suggests a 46% reduction in emissions from 1990 levels now looks likely.  This, while a somewhat smaller decrease than shown in these scenarios, would nevertheless likely result in emissions below the cap throughout the 2020s.

Chart 1: Projected EUETS emissions under three scenarios compared with the cap

Source: Sandbag

With emissions so persistently below the cap the surplus, after decreasing to 2020, begins to grow again, and continues growing to 2030 (see Chart 2).  It does so despite the operation of the MSR.

Chart 2: Projected cumulative surplus under three scenarios

Source: Sandbag

With such a large and persistent surplus there is a clear risk of prices weakening. This is especially the case later in the decade, where reductions in coal use in power generation seem likely to reduce the need for generators to buy emissions as a hedge to cover forward contracts, which may in turn further reduce demand for allowances.

The problem of the chronic surplus arises because the cap is both undemanding and rigid. There are at present no mechanisms for automatically resetting it, and no measures such as price containment which might limit how low prices could go.

The best way to deal with this problem is simply to reduce the cap in around the middle of Phase 4. This would be in line with the principles of the Paris Agreement, which envisages signatories to the Agreement adjusting their commitments over time to bring them more into line with the agreed temperature targets.

Chart 3 shows the effect of resetting the cap in 2026 to match actual emissions.  Under the Base Case the surplus begins to reduce rapidly as a result of the cap being reset.  Such an approach could readily be made consistent with other reforms, such as introducing a price floor in the EUETS.

Chart 3: Effect on the surplus of reducing the cap in 2026 (Base Case)

Source: Sandbag

While the 2017 reforms to the EUETS were a major step forward they are unlikely to prove sufficient.  Further measures will be needed to make sure the EUETS is robust as emissions continue to fall.

Adam Whitmore – 9th April 2019

 

 

 

[i] With a 43% reduction from 2005 levels in the sectors covered by the EUETS.

[ii] https://sandbag.org.uk/wp-content/uploads/2019/03/Halfway-There-March-2019-Sandbag-3.pdf

 

Increasing the political acceptability of carbon taxes

Straightforward, practical measures can make carbon taxes more acceptable to voters.

Carbon pricing often faces political obstacles due to public opposition …

Carbon pricing has spread widely in recent years, with around 40 systems now in place[i].  However, most emissions are not yet priced, and, even where they are, most prices remain too low.

Both expanding coverage and increasing price levels face political obstacles.  Overcoming these is essential for carbon pricing to play the role that it should in reducing emissions.  Fortunately, evidence is now emerging on what can be done to reduce opposition from voters – overcoming opposition from powerful lobbies such as industry warrants separate approaches.

A study by researchers at the LSE’s Grantham Research Institute, based on reviewing 39 existing empirical analyses, describes people’s objections to carbon pricing and other kinds of environmental taxes, and suggests specific actions to overcome them.  (The study focusses on carbon taxes, and most evidence is from North West Europe and North America, so the conclusions may not extend fully to emissions trading systems or to other cultural contexts.)

The study identifies several reasons people oppose carbon taxes:

  • The personal and wider economic costs of a tax are seen as too high.
  • Carbon taxes are seen as regressive, having a disproportionately negative effect on low-income households.
  • Carbon taxes are not believed to be an effective way to reduce emissions.
  • Governments are seen as having a ‘hidden’ motive to increase fiscal revenue rather than curb emissions.

However the study noted that people’s aversion to carbon taxes decreases over time after they have been introduced, particularly if the effects of the tax are measured and communicated.

There are various design options for reducing public opposition …

The study then identifies a range of measures for addressing the objections

  • Phasing in carbon taxes over time, introducing the tax at a low rate but having commitment devices to subsequently increase the rate to more efficient levels.
  • Redistributing revenues to ameliorate the regressive effects of taxes.
  • Earmarking revenues for emission reduction projects, which is popular with voters and improves the perceived effectiveness of carbon taxes.
  • Ensuring revenue neutrality of carbon taxes.
  • In all cases, policymakers need to gather and communicate the objectives and design of the carbon price to improve trust and credibility, before and after the introduction of a carbon tax. This includes communicating emissions reductions achieved and co-benefits of reductions in other pollutants[ii].

Drawbacks to these options seem limited …

The study notes that these recommendations may diverge from “first best” tax designs recommended in the economics literature.  However, while the study does not assess the implications of this, it is not clear to me that, even where they exist, these divergences are very significant.  They seem to me likely to be easily outweighed by the increased acceptability (a “sub-optimal” carbon tax that can be implemented is usually better than an “optimal” one that can’t).  And there are likely to be benefits often omitted in modelling of “first best” designs. This is especially the case as once a tax is in place it can be modified to over time as experience is gained and acceptance increases.

For example, phasing in a carbon tax is likely to produce economic benefits by reducing economic dislocation due to a price shock from sudden introduction at its full level, which may at least partly counterbalance the inefficiencies from prices being below optimal levels for an initial period.  Similarly, redistribution of revenue to poorer households may provide an economic stimulus benefits as poorer households are more likely to spend the revenue than richer households.  It may also increase social solidarity in ways which are conducive to economic welfare and growth.

Other emissions reductions, for example improving building insulation and deploying new technologies, may be funded at more nearly optimal levels where there are currently restrictions.  However, caution is needed here, and there may often be a stronger case for dispersing funds to citizens.

Revenue neutrality can take different forms.  One approach is to use revenues to reduce other taxes.  This is the approach adopted for the introduction of the carbon tax in British Columbia.  Economists tend to favour this type of approach because existing taxes are seen as distortionary.  However this approach often lacks transparency and credibility even if accompanying tax cuts are publicised – for example if other taxes are reduced they may be increased again in future.  This appears to be one reason why voters tend not to prefer this option.

And the current Canadian experiment with “tax and dividend” approaches appears promising …

A stronger guarantee is provided when revenue is explicitly returned to citizens.  This approach is usually referred to as “tax and dividend” (or “fee and dividend”, or “cap and dividend” in the case of any emissions trading system).  I’ve previously noted the advantages of this approach (see here).  It has been implemented for the Swiss carbon tax in the form of rebates on health insurance costs.  Four provinces in Canada are now working on implementing dividends in the form of direct financial payments to citizens.  This will make most citizens better off as the result of the tax, because they will also benefit from revenue raised from businesses.

There is an argument made in the environmental economics literature that a lump-sum dispersal to citizens is economically suboptimal, because it is better to use funds to reduce other taxes and so reduce distortions.   There is little if any empirical support for this argument as far as I am aware.  But in any case taking a view that citizens have more of a natural claim on property rights to the atmosphere than governments makes the limitation of the argument clear.  From this perspective, not providing citizens with any of the proceeds from pricing emissions is in effect a 100% tax on those proceeds imposed on everyone.  This is indeed non-distortionary – it applies the same tax to everyone irrespective of circumstances – but a fixed per-capita tax is not regarded by governments or their citizens as a good idea anywhere, for sound reasons.

A larger objection to returning all revenue directly to citizens, or using it to reduce current taxes, is that emissions run down natural capital for the benefit of current generations at the expense of future generations.  Intergenerational justice would, as I’ve previously argued (see here and here), be better served by some combination of preserving natural capital and investing revenue from carbon pricing in a “carbon wealth fund” analogous to a sovereign wealth fund.  However this would be unlikely to increase the political acceptability of carbon pricing compared with immediate dispersal of revenues to citizens.

Overall, the study makes a range of recommendation that are well justified on a range of grounds, and seem likely to help establish carbon pricing more widely and effectively.  It is to be hoped that governments everywhere take note of the findings.

Adam Whitmore – 5th March 2019 

Thanks to Maria Carvalho for useful discussions about the background to the study covered by this post.

[i] See the World Bank’s State and Trends of Carbon Pricing report here.  The definition of carbon pricing adopted in that report is quite broad, but even excluding some of the systems included in the report there remain over 40.

[ii] Please see World Bank’s  Guide to Communicating Carbon Pricing here for more information on developing an effective communications strategy.

 

Rapid growth of solar PV seems likely to continue

A simple s-curve model of solar deployment shows continued strong growth.

In my previous post I looked at the IEA’s projections for solar PV.  These always project no growth (or even a reduction) in the rate of installation, whereas in practice the rate of installation keeps growing rapidly.  I commented in the post that the growth of solar deployment did not seem likely to stop any time soon.  So how fast might solar PV continue to grow?

To estimate how fast solar PV deployment will grow, I’ve adopted a simple logistic function (s-curve) model for the deployment of solar.  This type of function is widely used to model the growth of new technologies[i].  The results for two scenarios are shown in the chart below together with actual annual deployment to date.  Both scenarios fit the historical data well, and are similar for the next few years, but then diverge significantly.

Scenarios for deployment of solar PV

Source for historical data: BP statistical review of world Energy to 2017, estimate for 2018 based on data in previous post.

The low case is based on an electricity system continuing to grow at current rates, with solar taking an increasing share, and deployment eventually reaching 300GW p.a. (see notes below for more on this).  The base case assumes a larger role for the power sector in the energy mix, as decarbonisation drives the electrification of end use, and solar deployment eventually reaches 50% more than in the low case, at 450 GW p.a..

These projections show deployment in another 4 to 6 years reaching more than double its 2018 rate of just over 100GW.  This compares with the 3 years it took to double from 50GW to its present size.  By 2030 solar is generating 3600 to 4500TWh p.a., around 12-15% of electricity consumption[ii].

Of course this highly stylised analysis only gives an indication of scale, and even greater growth is possible.  However I have not included a higher scenario, as these scenarios already represent continued very rapid growth.  This will require continuing attention to how solar can best be integrated into wider energy systems, including through the greater use of battery storage.

Adam Whitmore – 6th February 2019

Notes: Developing indicative markers for eventual industry size

A low case is estimated by looking at the size of the power sector.  This requires (in very approximate numbers) about 1000TWh of new and replacement generation each year over the next couple of decades.  If a third of this were to be solar it would eventually grow to about 330TWh p.a. of this, or about 300GW p.a..  It seems unlikely that solar’s share of new capacity would in the long run be less than this given its cost competitiveness and scalability.

This scenario appears roughly in line with Shell’s Sky scenario.  Both suggest that by 2035 Solar PV generation will be a factor of a little over 20 higher than in 2015.

However, this eventual rate of deployment may be an underestimate.  Decarbonising the energy system will require widespread electrification of end use, and so much more of the world’s energy will come from low carbon electricity.  For this analysis I’ve chosen a figure for eventual installation rate 50% greater than in the low case, reaching of 450GW p.a..  This represents one possibility within the range of scenarios for more ambitious decarbonisation, and higher estimates are possible.

[i] A logistic function is often used to model deployment of new technologies based on a range of examples, and I’ve previously used this type of model to look at electric vehicle growth – see here including examples of previous technology transitions. The analysis presented here updates my previous analysis of solar in both data and approach, given the  additional data available since that was completed.

[ii] World electricity consumption was 21,000TWh in 2015, https://www.statista.com/statistics/280704/world-power-consumption/ growing at 2.6% p.a. over 2010 to 2015.  Assuming this growth rate is maintained electricity consumption will reach around 31,000TWh by 2030.  BP’s review of energy suggests a lower growth rate, with around 2000TWh less demand in 2030 than in the case used here, presumably reflecting greater efficiency.

The IEA’s solar PV projections are more misleading than ever

The IEA is still grossly underestimating solar PV in its modelling

This post is a quick update of previous analysis.

Back in 2013 I pointed out how far from reality the IEA’s projections of renewables deployment were.  They persistently showed the rates of installation of renewables staying roughly constant over the following 20 years at whatever level they had reached at the time of the projection being made.  In reality, rates of installation were growing strongly, and have continued to do so (see chart).  Rates of installation are now a factor of nearly four times greater than the IEA was projecting back in 2013 – they were projecting installation rates of about 28GW for 2018, where in fact around 100 GW were installed in 2017[1] and an estimated 110GW in 2018.

I have returned to the topic since 2013 (see links at the bottom of this post), as have many others, each time pointing out how divorced from reality the IEA’s projections are.

Unfortunately, the IEA is continuing with its approach, and continuing to grossly understate the prospects for renewables.  Auke Hoestra has recently updated his analysis of the IEA’s solar PV projections to take account of the latest (2018) World Energy Outlook New Policies Scenario (see link below chart – in addition to chart data his post also contains a valuable commentary on the issue).  The analysis continues to show the same pattern of obviously misleading projections, with the IEA showing the rate of solar PV installation declining from today’s rate until 2040.  Of course eventually the market will mature, and rates of installation will stabilise, but this seems a long way off yet.

IEA projections for solar PV in successive World Energy Outlooks compared with outturn

http://zenmo.com/photovoltaic-growth-reality-versus-projections-of-the-international-energy-agency-with-2018-update/

In 2013 I was inclined to give the IEA the benefit of the doubt, suggesting organisational conservatism led to the IEA missing a trend.  This no longer seems tenable – the disconnect between projections and reality has been too stark for too long.  Instead, continuing to present such projections is clearly a deliberate choice.

As Hoekstra notes, explanations for the disconnect have been advanced by the IEA, but they are unsatisfactory.  And as renewables become an ever-larger part of the energy mix the distortions introduced by this persistence in misleading analysis become ever greater.

There is no excuse for the IEA persisting with such projections, and none for policy makers taking them seriously.  This is disappointing when meaningful analysis of the energy transition is ever more necessary.

Adam Whitmore -21st January 2019

https://onclimatechangepolicydotorg.wordpress.com/2013/10/08/why-have-the-ieas-projections-of-renewables-growth-been-so-much-lower-than-the-out-turn/

https://onclimatechangepolicydotorg.wordpress.com/2015/02/27/the-ieas-central-projections-for-renewables-continue-to-look-way-too-low/

https://onclimatechangepolicydotorg.wordpress.com/2015/06/27/the-ieas-bridge-scenario-to-a-low-carbon-world-again-underestimates-the-role-of-renewables/

https://onclimatechangepolicydotorg.wordpress.com/2017/09/26/underestimating-the-contribution-of-solar-pv-risks-damaging-policy-making/

[1] The BP Statistical Review of World Energy shows a total of 87GW installed in 2017 https://www.bp.com/content/dam/bp/business-sites/en/global/corporate/pdfs/energy-economics/statistical-review/bp-stats-review-2018-renewable-energy.pdf