Monthly Archives: October 2013

Can the EUETS combine intensity-based and absolute emissions caps?

An innovative reserve mechanism for the EUETS is being considered that moves allowances to and from the reserve based on the level of economic activity.  This would give the EUETS elements of an intensity based cap (limiting emissions per unit of output) within an absolute cap (limiting total tonnes emitted).   Although potentially less economically efficient than managing reserves based on prevailing prices it may prove politically more tractable. 

The prevailing surplus of allowances in the EUETS is leading to the scheme lacking effectiveness as a signal for abatement, and especially for low carbon investment.  The EU is considering a range of reform options to address this.  One option is to cancel allowances currently due to be auctioned, although this is likely to face substantial opposition.   Another option, not mutually exclusive, is to establish a reserve of allowances to stabilise the market.  Options for the reserve now under discussion are all volume based, as they seek to stabilise the market by using a reserve to manage the volume of available allowances.  But they differ in the basis for the trigger mechanism used to determine the timing and number of allowances transferred to or from the reserve.  There are three main types of basis for a trigger mechanism: volume of allowances, levels of economic activity, such as GDP, and allowance prices.  I assume here that the mechanisms applies automatically, but it would be possible for it to be administered on a more discretionary basis, analogously to the types of functions performed by a central bank.

I have previously written about the advantages of allowance reserves that use price based triggers to implement soft floors and ceilings on the price, such as those already implemented in California.  However, even if such mechanisms represent the best way forward in principle they may prove politically impossible to introduce in Europe at the moment.  So what about the other two possible types of trigger?

A volume based trigger in which allowances are put into the reserve when the cumulative surplus exceeds a specified upper threshold, and moved back into the market when the surplus (although still present) falls below a lower threshold, is essentially a form of temporary set-aside.  If allowances thus automatically come back into the market as it tightens, the timing and extent of the market’s return to cumulative scarcity appears likely to remain largely unaffected.  Market participants will anticipate the return of allowances to the market and factor it into their pricing.  Such a mechanism thus seems unlikely to add significantly to market stablility.

An economic activity based trigger is quite different.  It seeks to make the supply-demand balance more stable by adjusting the supply of allowances in response to the level of economic activity.  Allowances are placed into the reserve if economic activity is lower than expected and withdrawn from the reserve if activity is higher. This can result in supply being reduced into the long term after a recession, because allowances only come back into the market when a period of higher than expected economic growth leads to economic activity above expected levels.  This may not occur, or may return only a proportion of allowances.  It turns the scheme into something more like an intensity based scheme, where emissions per unit of activity are limited, but in this case still subject to an overall cap.

Such a mechanism does not address all of the causes of over or under supply.  Consequently it may not prevent price falling very low or rising very high in some circumstances, and so inefficiently low or high prices remain a possibility.  However a price-based trigger would probably be needed to avoid such risks completely.

This type of mechanism requires the measure of economic activity used as a basis for the trigger to be defined.  This may be GDP, which measures (however imperfectly) activity in the economy as a whole, or it may be something that represents activity in the sectors covered by the EUETS, for example a mix of electricity consumption and industrial output.

The expected level of economic activity that corresponds to no transfers to and from the reserve needs to be set.  It will likely be appropriate to reset this from time to time to take account of changed expectations, at least at the start of a new phase of the scheme.

The number of allowances that are transferred to or from the reserve in response to changes in the level of economic activity (elasticity) also needs to be defined.  For example if economic activity is 1% lower than expected then 1% of the cap may be put into the reserve, or some other proportion such as 0.75%.  The response may include limits on the number of allowances transferred in any one period, for example a quarter of auctioned volumes in any one year.  And it would be possible to specify no transfer in the event of small changes in economic activity relative to the expected level.

The way that such a mechanism might have worked had it been in place since the start of Phase 2 is illustrated in the chart.  For simplicity it uses GDP as an indicator of economic activity.  Expected GDP at the start of Phase 2 in 2008 was 2.3% p.a, with rising economic activity over time (dashed blue line).  Actual GDP has been well below this (solid blue line), and is currently around 12% below expected levels.  This leads to allowances being put into the reserve (green bars and lines which show annual and cumulative totals, assuming an elasticity of 0.75).  This is close to the cumulative surplus due to lower emissions shown on the chart (dotted black line).  The surplus is defined as the difference between the annual cap and actual emissions (dashed and solid grey lines), and excludes any surplus due to other factors such as the use of offsets.  In a hypothetical future the reserve continues to grow as GDP remains below expected values.  Then, after expected GDP is reset at the start of Phase 4 in 2021, a period of more rapid than expected growth begins to reduce the size of the reserve.  However a large volume allowances remains in the reserve even though the cap has significantly tightened and the market is likely to have returned to scarcity.

reserves chart corrected v2

It is not clear whether or not such a reserve mechanism, in effect providing an intensity based limit subject to an overall absolute cap, is a good idea.  It may fail to meet policy objectives in some circumstances, and other options for reform may be preferable.  But it seems worth further consideration.  If nothing else it may at least help remove at least some allowances from the market if permanent set-aside (cancellation) of allowances is not politically feasible.  And as an innovation among emissions trading schemes it could go some way to restoring the reputation of the EUETS, and would provide a signal to others around the world that the EU is willing to take action to address the problems with the EUETS as it now stands.

Adam Whitmore – 24th October 2013

To follow this blog click on the button in the bottom right hand corner of the screen.

Why have the IEA’s projections of renewables growth been so much lower than the out-turn?

The IEA has greatly underestimated the growth of renewables for some years now.  This illustrates how important it is to allow for unexpected outcomes if policy design is to be robust, as even well informed projections can be very different from the subsequent out-turn.

(For an update on the IEA’s projections of renewables see also this post.)

The International Energy Agency’s (IEA’s) annual World Energy Outlook (WEO) is a thorough and well researched analysis of the outlook for the world’s energy systems[1].  Over the years it has become the standard view of the world’s energy use now and in the coming decades.  However it has had an extraordinarily poor track record in projecting the growth of solar and wind power in recent years.  The charts below compare the IEA’s projections over the last few years with the out-turn for both wind and solar.  Projections have been revised upwards each year.  But they have still been consistently too low, by a very large amount in most instances, with the pattern persisting over many years for two different groups of technologies, wind and solar PV.   As recently as 2006 it was expected to take until the 2020s to reach current levels of wind capacity, and until the 2030s to reach current levels of solar capacity, with current solar PV capacity almost an order of magnitude greater than expected in just seven years ago.

The IEA’s projections have consistently increased over the years, but still fallen short of actual deployment ….

wind and solar past projections

It would, of course, be wrong to suggest that because past projections have been underestimates the current projections will also be too low.  However the most recent projections continue to show rates of deployment that appear very cautious.  The graphs below show the IEA’s projected rate of installation in the most recent WEO (for 2012) under its central New Policies scenario compared with past and current growth rates.  For both wind and solar projected installation rates start below 2012 levels and remain roughly constant or fall over time.

IEA’s projection show declining rates of deployment for both wind and solar …

wind and solar current projections

Decreases in installation rates are of course possible.  Wind installation seems likely to be lower this year than last, although the rate of solar deployment continues to grow.  However, projecting flat or slowly declining installation rates over the next couple of decades suggests either that current rates are a spike, or that installation is moving towards saturation.  Neither of these possibilities seems likely.  Costs are continuing to fall, especially for solar, renewables still account for a very small share of total generation, and drivers towards deployment of low carbon technologies seem likely to strengthen rather than weaken over the period.  One does not need to be an advocate of renewables to expect that these industries are more likely to grow than shrink over the next couple of decades, even if growth of annual deployment may be much slower than in the past.  It would seem more plausible if a central case scenario were projecting some continuing growth in annual installation, with decreases very much a low case.   It will be interesting to see how these projections are adjusted in the next edition of the WEO due out in a few weeks.

So what has led to this persistent underestimation of growth?  There may have been a reliance on individual jurisdictions’ plans, with more caution than seems with hindsight to have been warranted about the rate at which policy might move.  This seems to have led to linear extrapolation of capacities when technologies were in a phase of exponential growth.  Projections for wind have improved in recently years as growth appears to have become more linear (at least temporarily), and following a large upward revision in the projected rate of addition between the 2009 and 2010 editions of the WEO.  It may also be that there is some inherent caution about new technologies.  However the IEA – along with many others – has tended, if anything, to be somewhat optimistic about CCS, so this cannot be a complete explanation.  There are also specific circumstances that have played a role, notably being somewhat slow to recognise the falling costs of solar PV, with even the costs from the 2012 edition being well above actual values[2].

There may also be a deeper explanation rooted in institutional conservatism.  Taking a conservative view of future prospects in the energy sector can be necessary to avoid being swayed by the latest fad.  A conservative view recognises the realities of the long time horizons and vast scale of the world’s energy systems.  However it can carry the risk of missing the role of genuinely transformative technologies, as appears to be the case here.  The IEA’s current caution may still prove justified.  But  Eurelectric, the European power industry association, noted in a recent report that the European power sector is already undergoing one of the largest transformations in its history[3].  Such changes seem likely to be a global phenomenon.  Wind and (especially) solar PV seem likely to form part of the largest transformation of the energy sector at least since the growth of oil consumption in the middle decades of the 20th century, and perhaps since the invention of the steam engine.  The IEA seems to be slow to recognise this.

Whichever way the future turns out, the IEA’s past projections show how different actual out-turns can be from even well-informed projections.  This provides and important reminder that none of us can be sure about future changes to the energy sector, and policy design must always be robust against things turning out to be different from expectations.

Adam Whitmore – 8th October 2013

ETS price floors and ceilings

A more up-to-date review of this topic can be found on the price floors and ceilings page of this site – page number 3 under the Carbon Pricing tab above, which can also be reached via this link:

Mechanisms that limit prices under an ETS (price floors and price ceilings) are already in operation in North America.  Similar arrangements can easily be adopted elsewhere, including in the EU.

In previous posts I have alluded to the benefits for an ETS of price floors in the form of auction reserves and soft ceilings in the form of a price containment reserve taken from within the cap.  In this post I take a look at how such arrangements work in North America to demonstrate that they are not esoteric theoretical ideas, but existing practice that can easily be adopted in similar form in the EU or elsewhere.  California, Quebec and the Regional Greenhouse Gas Initiative (RGGI) have a soft price floor in the form of an auction reserve price.  Prices can fall below this floor, but if bids are below this level in the auction allowances are not sold, so supply quickly tightens to maintain the price at the floor.  When prices rise, allowances can be released from a reserve acting reducing upward price pressure, although California and RGGI differ in the extent to which allowances additional to the cap can be offered into the market.   They also have mechanisms to limit price rises in the form of allowance reserves, but differ in the extent to which additional allowances can enter the market.

Price floors

Under the California ETS (Quebec has very similar arrangements) there is an auction reserve price of $10/tonne in 2012, which rises at 5% p.a. plus an inflation adjustment, and is currently $10.71/tonne.  Any allowances that are not sold at auction are retained by the regulator, the Air Resources Board (ARB), in an Auction Holding Account. The holding account allowances are not made available again through the auction until the price has exceeded the floor price for two consecutive quarterly auctions, and return is subject to a limit of 25% of the total allowances available at each regular quarterly auction. As a result, a surplus in the Auction Holding Account may take time to be drawn down.   In practice all 2013 allowances were sold out at the first four auctions, and the market price is above the floor level, though only by a little in the first auction.

RGGI, which covers power sector emissions from several states in the north eastern USA, also includes an auction reserve price.  However, the reserve price is much lower than in California, at $2/short ton in 2014 rising at 2.5% p.a., approximately in line with inflation.  In the past the floor has often been binding, but the current price is above the floor.  Allowances unsold at auction prior to 2014 are retained by the authorities and can be auctioned again, but allowances unsold at the end of each 3 year control period (the current control period is 2012-2014) may be retired permanently at the discretion of individual states.  This gives a possible mechanism for automatically tightening the cap if there is a surplus allowances at the floor price over an extended period.

Price ceilings

In California there is an Allowance Price Containment Reserve (APCR) from which allowances are released at prices of $40, $45, and $50/tonne in 2013, rising at 5% p.a. plus inflation thereafter.  This is entirely separate from the Auction Holding Account used for the floor.  Allowances are sold from the APCR on a quarterly basis if there is demand.  The sale is held six weeks after the regular quarterly auction of allowances, allowing buyers to make up a shortfall after the auction. Buyers specify the number of allowances they want at any of the three fixed prices.

122 million allowances have been put into the APCR for the period to 2020 equal to 4.5% of the overall cap across all years (including the maximum allowed offsets), and relative to a single year (2015) is 29% of the cap including the maximum allowed offsets.  The APCR allowances are taken from within each year’s capped total.  The reserve is divided equally among the three price tiers.   ARB has proposed that the number of allowances in the APCR be increased to 207 million tonnes, which would be 7.6% of the overall cap.

RGGI also has a costs containment reserve (CCR) of additional allowances that can be released into the auction when the auction clearing price crosses a certain threshold.  As with the floor, the prices at which allowances are released are much lower than in California, being $4/short ton in 2014 rising at $2/short ton p.a. to reach $10/short ton in 2017, escalating at 2.5% p.a. thereafter. The CCR allowances are in addition to the cap, and balances are re-set annually to 10 million tons (which is just over 10 percent of the 2014 cap) if allowances are drawn down from the CCR.  Unlike California, the cap is thus effectively loosened if there is continuing demand for allowances from the reserve, and so functions much more like an absolute ceiling.  (These are the revised the rules for RGGI to apply from 2014 forward, which has yet to be fully enacted by all individual states, but I understand that adoption is expected to be completed shortly.)

Similar arrangements could readily be made for the EUETS.  A price containment reserve of the same proportion of the 1013-2020 cap as Californian (4.5-7.6% of the total cap for Phase 3 excluding aviation) would be approximately 690 to 1170 million allowances, very much in line with the quantities that have been discussed for set-aside or backloading.  A market stabilisation reserve would be, in effect, somewhere between backloading and permanent set aside.  This is because allowances placed into the stabilisation reserve could potentially return to the market (in contrast to permanent set-aside).  However, they would do so only in response to prices reaching threshold levels, rather than automatically (as is intended with backloading).  This type of reserve could be accompanied by an auction reserve price to stabilise the price at the lower end of the range.

Some in the EU continue to oppose price containment mechanisms.  Whatever objections may be advanced against price containment – and none seem compelling to me in principle, at least for an auction reserve price – a lack of practical examples elsewhere on which to draw is certainly not one of them.

Adam Whitmore –  2nd October 2013