Why Norwegian electricity should not be considered green outside Norway

12 min readAug 4, 2019

TLDR: Norway (and Iceland, and Denmark) produce 100% (Denmark: 70%) of their electricity from green sources. They however sell most of their GOs (certificates of origin) to other countries, supposedly because they don’t need to use them themselves because everyone knows about their green credentials.

This in effect double counts their certificates: once implicitly (“everybody knows that Norway only has green energy”) and once explicitly in another country using the the GO. As we can’t do much about the implicit counting we have decided that we will not use their certificates for other than the local markets.

At Green Hallmark we have recently been asked recently why we do not recommend Norwegian GOs despite 100% of the energy in Norway being produced from renewable sources, and Norway being one of the greenest countries in the world. The reason in a nutshell is simple: Norwegians do sell a lot of their GOs to others, and have actually a rather dirty residual mix once you have taken this into account. On the other hand they supply a lot of GOs to the European market which in turn suppresses the prices and therefore does not particularly support the investment of green energy installations.

We believe that if you want a working EECS system then you need to exclude most Norwegian GOs from the other European markets, with the exception of new capacity that is being brought online to supply the energy needs of Europe. We will go of this argument in more detail below.

The EECS GO system

EECS stands for European Energy Certificate System and it is the system that deals with creating, distributing, and retiring Guarantees of Origin (“GOs”) across the European Union and some neighbouring states. We’ll discuss how it works in more detail below, but let’s first look at the economics behind GOs in general.

The mechanics of green energy GOs

Taking a step back, what we want to achieve is that end-customers (private or commercial) who care about their green credentials can ensure that they only use electricity coming from green sources. In fact — they’d actually like to be able to pinpoint the exact source where their green electricity is coming from to allow for a genuine choice, and avoid green-washing by an overly loose certification system. Unfortunately electricity — especially green electricity — has some properties that make this a bit complex

electricity can not efficiently be stored, so it should be consumed at the time it is produced, regardless of the demand schedule of the consumer and the supply schedule of the producer
electricity can be transported over long distance, but the infrastructure needed to do this is (a) limited in capacity, (b) heavy from and environmental and landscape point of view, and © suffers some transmission losses
electricity can not be individually routed, and all consumers online at one point in time de facto jointly consume the electricity mix provided by all producers online at the same time.

The points above together make it impossible for consumers to physically choose where their energy is coming from, and therefore make impossible for them to be physically supplied from the plant of their choice (as an aside: consumers can of course have their own, co-located production plants; however, this only works insofar as there is a perfect co-incidence of supply and demand; in practice even those customers will have to either deliver excess electricity into the network, or consume electricity from the network, or both, depending on the time).

The way we get around this problem is by separating the physical electricity from its green credentials, and to allow for a separate distribution of the green credentials and the electricity they relate to. They way it works in practice is that every producer of green electricity within the scope of the EECS gets a GO for each and every MWh of green electricity they produce. Those GOs provide plant-level transparency on the origin of the electricity, and also certify certain characteristics of the plant (eg, type, year of construction, location) as well as the month and year to which the certificate relates to.

GOs are freely transferable and negotiable, so they can be traded until they reach the end-customer who wants to actually use them to “upgrade” the “grey electricity” they used (all electricity drawn from the network is considered “grey”) into “green electricity”. The process of using the certificates is referred to as “retiring” them, and once they are retired the beneficial owner gets a “retirement certificate” and the corresponding GO can no longer be transferred.

The economics of green energy GOs

Taking a step back it is important to understand the logic behind the EECS GO system before we can move on to the economics. They key assumptions of the system are

more and more customers care about where the electricity they are using is coming from, and they are willing to pay a premium for sustainably sourced electricity
for reasons related to the physics of electricity, the only way for at scale giving customers the ability to choose the origin of their electricity is by separating the green credentials from the “grey” electricity at the point of production

For this system to work it is essential that whoever uses a GO to upgrade their grey energy considered having used green energy from the particular plant that this GO relates to, and it is even more essential that noone else — in particular not those who have consumed the actual energy produced by the plant — can claim having using used it, otherwise we have double counting of the green credentials.

Why is this important? Let’s get back to why we have GOs: we want to provide plant-level transparency of energy consumption, allowing consumers to choose from where the consume their energy, and thereby channel funds to the most sustainable electricity production facilities. This however is not and end in itself: the underlying goal is to make the development of genuinely green electricity plants more attractive and therefore encourages them being built. This in turn means we need to be a bit careful which certificates we want to recommend, because whilst the 50 year-old dam built with a massive impact on the local eco-system might be (and arguably should be) able to issue GOs just like the newly-build solar plant next door, but if I want to have an impact I want to support the latter (and, to the extent that I can, future developments) rather than providing the former with a windfall profit.

The residual mix

Before we can go on we have to briefly deal with the “residual mix” which corresponds to the electricity that consumers of grey electricity get. It is essentially the electricity produced in the system, minus the green electricity that has been explicitly transferred out of the grey electricity pool by means of GOs. So importantly the residual mix can contain green energy to the extent that the corresponding green certificates have not been used.

To explain this we’ll go through and example calculation. Note that technically the residual mix can only be asserted with a year delay because only at this point it is certain how many GOs have been used (unused GOs will have expired by then) but we’ll ignore this complication and assume here for simplicity that GOs are used immediately (or, more realistically, a steady state).

To have round numbers, let’s assume a (small) country with 100 GWh of daily production and consumption, and let’s assume 30% are fossil, 40% nuclear, the remaining 30% are green and have the right to GOs. If all GOs are used locally then the residual production is 30 GWh fossil and 40 GWh nuclear, so the residual mix is 43% fossil and 57% nuclear, and the percentage of green energy usage is 30%. If however only 20 GWh of GOs are actually used then the remaining 10 GWh of green energy stay in the residual mix which then becomes 37.5% fossil, 50% nuclear and 12.5% green. In this case the green energy usage is still 30% (30 GWh), out of which 20% come from the purchase of GOs, and the remaining 10% (10 GWh) come through the residual mix.

Things get slightly more complex when this country exports or imports energy and/or GOs. If a country exports electricity (but not GOs) then they export their residual mix. This does not impact their residual mix, but it does impact the proportion of green vs grey energy used. In the above example, and assuming the 20 GWh usage scenario, if the country exports 20 GWh of grey energy the residual mix is still 37.5 / 50 / 12.5. However, the country’s green energy usage is now 20 GWh (from the GOs) plus 10 GWh * 50/70 = 7.1 GWh (from the non-exported portion of the grey energy mix), so the country’s green energy usage is now 27.1 GWh / 80 GWh = 33% which is higher than the initial 30% because grey energy was exported but not the corresponding green certificates.

Things are similar when the country exports GOs but it does not export the corresponding electricity, except for the twist that in this case if effectively imports the residual mix (after applying the purchased GOs) from the country to where it exports the GOs. Lets assume a country A whose production mix is 70% fossil / 30% nuclear and a country B whose mix is 100% green, both producing and consuming 100 GWh per day. Let’s further say that country A purchases 45% of country B’s GOs, ie 45 GWh. In this case country A’s residual energy consumption is 55 GWh with a residual mix of 70% (38.5 GWh) fossil / 30% (16.5 GWh) nuclear, and its percentage of green energy usage is 45%. Country B in this case uses 100 GWh of electricity but only 55 GWh of green electricity because it sold the 45 GWh of GOs to country A. Those 45 GWh replaced the same amount of grey energy at a residual mix of 70 / 30, and to keep numbers consistent at the overall level this is the mix to be applied to country B. Therefore country B, after taking into account the GOs sold, has a residual mix containing 31.5 GWh fossil and 13.5 nuclear. The actual residual mix of country B depends on how many GOs are used internally: if no GOs are used in country B then the residual mix is 31.5% fossil, 13.5% nuclear, and 55% green; if 55 GWh of GOs are used then the residual mix is 70% fossil / 30% nuclear. In both cases the green energy usage in country B is 55%.

Now all of this is a bit complicated but the underlying principle is easy: if a country exports GOs without exporting the associated grey energy then it effectively imports the residual mix of the country to where it exports the GOs.

The European GO market

After we went through all the theory, let’s now look at the European GO market in practice. The first chart below is a country’s residual mix, ie the energy mix of the grey energy (ie energy that has been purchased without using GOs) on that particular market.

In the second chart below, each country has two bars. The left bar is the production mix, and the right bar is the supplier mix (aka consumption mix), ie the mix after application of GOs that the entirety of consumers of this country has purchased. As opposed to the residual mix above that only covers the grey energy sold on the market, the supplier mix covers all energy consumed in that market.

Note that the chart above is normalised to 100% production and consumption respectively, which may or may not be the same amount. This for example explains the percentage of nuclear energy in Luxembourg, because Luxembourg imports a lot of their energy from France, and therefore a lot of the energy imported will be nuclear, despite Luxembourg not having any nuclear plants.

So now let’s look at Norway. Virtually all of their energy production is green, with only a very small portion being fossil, and no nuclear. However, in the supplier mix, only about 50% of Norway’s electricity is green, the remainder being mostly fossil and a bit of nuclear. The reason for this is along the lines that explained above: Norway exports a significant proportion of their GOs, and therefore it imports the residual mix of the countries to where the GO’s are exported. Now at 50% of the energy usage Norway is still not bad in the European comparison, but it is at best in the midfield, and for example behind countries like Switzerland.

Other countries in a similar situation are for example Denmark, which produces around 70% of the electricity from green energy sources, but whose supplier mix is only about 25% green. Croatia is also in this group of countries, albeit in much smaller proportions: their production is 75% green, however on the supplier side it is down to 55%. Last but not least, the most blatant case here is that of Iceland. Their energy production is 100% green, however, the final supplier mix only contains 22% of green energy. On top of this, Iceland, due to its geographic situation, has no interconnectors to any other regional electricity grid.

Conclusions and recommendations

Conclusions

The fundamental policy conclusion is that if we want to use the EECS to support the changeover to green energy we have to take it seriously. This in particular implies that in terms of assessing whether or not a product or service has been provided with green energy what matters is the energy mix post-application of all GOs and the production mix is irrelevant, regardless of the routing of the actual electricity.

A couple of conclusions from this fundamental assumption:

Despite having one of the highest green energy production rates of about 98%, at only 52% green energy in the supplier mix Norway is at best in the mid-field of European energy consumer countries, behind countries like Switzerland, Austria, Sweden and even Germany; the residual mix in Norway is relatively heavy in green energy at 43% (suggesting no-one bothers to buy GOs locally), and this is the level where generic Norwegian consumption should be pegged. In particular, statements like “Norwegian electric cars run 100% on green energy” should be considered false unless they use GOs. Instead one should assume that Norwegian cars, like all other Norwegian grey electricity consumption, run at 42% green energy.

A similar conclusion should be reached for Denmark, except that numbers are much worse: despite almost 70% of the energy produced there is from green sources the overall supplier mix is only about 25% green, and the residual mix in Denmark is only 10% green. So if you fill up your electric car in Denmark using grey energy it will be 70% fossil, 20% nuclear, and only 10% green.

The worst discrepancy is of course Iceland where 100% of the energy produced is from green sources, but only 25% in the supplier mix is, and <5% in the residual mix. So some consumers (20–25% by volume) in Iceland care about green energy which is the good news. The bad news is that 75–80% don’t. Our assumption is that the latter are mostly the aluminium producers, and possibly the bitcoin miners. The conclusion in this case should be that unless you purchase aluminium produced in Iceland (or “bitcoins mined in Iceland”) with a green certificate attached you should assume the residual mix of 60% fossil, 35% nuclear, and 5% green.

Germany also has an interesting situation. The production mix is about 37% green (and 11% nuclear, so with a massive fossil component), but the supplier mix is 55% green. Moreover the residual mix is <1% green. This suggests that Germans are very diligent with purchasing their green certificates. However, those that do not purchase them, effectively consume 80% fossil and 20% nuclear which is amongst the worst residual mixes around (of the big countries, only UK and Netherlands are in a similar situation).

Last but not least, special mention to Poland and, to a lesser extent, Italy: they already have a relatively low percentage of renewable energy production (12% and 40% respectively) but they also export the large majority of them, leaving the residual mix at 7% and 22% green respectively.

Policy recommendations

From the above we derive the following policy recommendation

Unless you are located in Iceland, do not purchase Icelandic GOs; also, if you buy Icelandic products (notably aluminium), unless they explicitly state they have been produced with green energy, then assume a mix of 5% green, 35% nuclear, and 60% fossil
A similar reasoning applies to Norway and Denmark, even though with a better residual mix to be applied to locally produced products (Denmark 10% green, Norway 50% green); here we also recommend to only purchase GOs for use in the respective country, with one exception: new plants that do not replace old green plants (and therefore export their energy) can be supported on a case-by-case basis
When purchasing energy-intensive products from Germany, UK, Netherlands and Poland ensure that they been produced using green energy, because otherwise their supply mix (based on those countries’ residual balance) is very bad.

Those are the most important policy recommendations to be drawn from the above, but of course we at Green Hallmark go to a much more granular level when we recommend which green energy a certain consumer should use, also based on grid transmission capacities.

***

Stefan Loesch is co-founder of Green Hallmark, a company whose mission is to encourage the production genuinely green energy in Europe and the world.

All data from “European Residual Mixes”, Version 1.2, 2019-07-11, published by AIB (see here)