Power - and power prices - are a hot topic in Norway. Questions like “should customers be exposed to market prices?”, “Does the market actually work at all?”, “Why do we pay so much for our electricity?”, “Why should we export to our neighbours?” Have reverberated through the echo-chambers of social media, and have been (largely poorly) addressed by more mainstream media outlets over the past few months.
Here I’m not going to discuss the ins and outs of the “Norgespris”, or whether we should have more or fewer cables to neighbouring countries. What I will do though is try to answer one of the key questions that keeps popping up in this debate, especially amongst non-industry people: why do we pay “a lot” for power in Norway when 1) our production is basically hydro, and 2) hydro costs almost nothing to produce?
It seems, on the face of it, that someone’s having a laugh - and running off with all our money. If hydro costs a fraction of an ore per kWh to produce, why do we pay man 10x that for power? Are the power elite somehow running off with all our money for expensive holidays in the Caribbean? The truth however is much more interesting than a conspiracy theory - and (unfortunately) also more complicated.
FACT BOX: the difference between a MW (megawatt) and a MWh (megawatt hour). “MW” measures the capacity of a power plant to produce electricity (its “power”). The more capacity (the more MW), the more energy a power plant can supply in a given unit of time. Megawatt hours, MWh, measures the amount of energy produced. Simply put, a power plant with 1MW of capacity can produce 1 MWh of energy in an hour running at 100%. Both numbers are important. Let’s say I am cooking an egg on my induction stove top. The cooking element may have a wattage of around 2000 W - meaning it draws 2 kW (=0.002 MW) of power at any given instant of time. If it takes me 15 minutes to boil the water and cook my egg, I will use 2 kW * 15/60 = 0,5 kWh of energy. I need the power system to deliver both of these - 2 kW of power and 0,5 kWh of energy over a 15 minute period.
First, some facts and data. What does the power sector look like in Norway? We have - very roughly - around 34 GW (that’s 34 million kW) of hydro generation capacity in Norway, and around 87 TWh (that’s 87 billion kWh) of reservoir storage capacity. Hydro produces around 135 TWh in a normal (“average”) year. Norway consumes a little under 130 TWh per year, and the year’s peak demand is around 24-25 GW. This (net) difference between production and consumption is exported.
The first thing that should stand out is that the storage is quite a bit less than the production (and consumption). At most, we can store around 2/3rds of our annual production if (and that’s a big if - so big its not practicable) every singe reservoir was 100% full at the same time. Given that we can’t store all we need in a year, how ensure we can meet demand for power when it is demanded? Inflow into the reservoirs happens throughout the year, and production occurs throughout the year. So, if we can manage both parts - ensuring we have enough capacity to store incoming inflows, and can produce the power we need when we need it - we can achieve this without having to store a full year’s worth of power at once.
However, this is not easy. Firstly, much of the inflow to these reservoirs comes in spring, when the snow melts. If the reservoirs are too full at the start of spring, we will end up not being able to store the inflow and spilling it into the sea. It should be obvious that this is wasteful - it would mean we are throwing away a resource that could be used to power Norway’s economy. But we can’t be completely empty at the start of spring, because we don’t know exactly when the spring flood (“flom”) will occur - maybe in March, maybe in April, maybe in May, or even June. And we have to ensure we have enough water stored to produce power in the meantime. So, that’s planning problem number 1: how to ensure we have enough water in storage at the start of spring to cover us until the spring flood inflows, but not too much so that we spill it all into the sea? Problem 1b is: how do we then manage this storage through the rest of the year?
The second big planning problem is a result of the fact that “Norwegian Hydro” is not a single power station, but thousands of reservoirs, dams and generators, located all over the country. Why is this a problem? Because if we manage the production badly - that is produce too much from some power plants and too little from others, those we have produced too much from will run out of water. Similarly, those we have produced too little from may end up with too much water to store, and have to spill. Spill, as we have seen, is bad for Norway AS (it’s the same as burning money). And the problem of running out of water in some plants is that you may not have enough generation capacity (that is, MW) to meet demand in peak periods.
What do I mean by the last point? Let’s assume we have a simple power system, where we have two hydro plants with 1000 MW of generation capacity and some storage. Let’s assume our demand maxes out at 1500 MW at breakfast time on a cold winters day. If both plants have water in their reservoirs, all is fine and dandy - both produce (let’s say 750 MW each) and the consumers make their breakfast, heat their homes and run their businesses. But what happens if we were rubbish planners and had over-used one of the plants over the previous summer and autumn, and it had no more water in its reservoir? Then we would only have 1000 MW to cover the 1500 MW of demand. This means only one thing. Load shedding and blackouts. A closed factory, a cold house, and no poached eggs for breakfast for you, consumer.
What’s more, the plant that we didn’t use may have ended up having to spill, and now the total amount of water in storage at that plant - even though the reservoir is full - may not enough to get us through the winter. So we run out of energy (MWh) as well. Oh well, glad I have a wood oven.
So, we see we have to be careful about when we use power from each of our power stations, to ensure we can meet both our total energy use over the year, and our demand at any given point in time. This is a real challenge. Norway has ostensibly 34 GW of production capacity, and peak consumption is around 25 GW. But in practice there is never a point where all of the production capacity is available at the same time. Some of the stations will be out of operation for maintenance or failure (this cannot be avoided). Some of it - especially the small local “run-of-river” stations may have very low river flows at certain times of the year and hence be unable to produce at maximum. And there is grid loss - not everything that exits the power plant enters our homes - some of the energy is lost on the way (that’s physics, baby). In total, it means that at best Norway has only a few GW of production capacity more than we need in our peak periods. And in dry years, where some plants have essentially run out of water (because there just isn’t enough, even with the best planning), we may not have enough.
The third big planning problem is that, even though they may be very smart, people in the power industry cannot for see the future. We don’t know how much it’s going to rain in 2025. We don’t know when the spring floods will happen and how much water will come. We don’t know exactly how much will be demanded by consumers and when. We can guess, and using data, models and analysis make very very good guesses, but there is a huge amount of uncertainty. And we still have to make good plans even with this uncertainty so that Norwegians have the power they need, when they need it.
So, we have these three big planning problems (there are more, but that’s enough for now):
1. Storage is limited. Usage needs to be timed right so there is always production available, and that we don’t tip money down the drain by spilling water.
2. There are thousands of hydro reservoirs and power plants. If we don’t coordinate well, we will end up using too much from some and not enough from others, and a. won’t be able to meet peak demand, and b. may even have to spill and at the same time not have enough in storage overall.
3. We don’t know what is going to happen, but need to plan anyway.
So, given this, how do we plan? Let’s look at how it is done today. Each power company does this independently from each other, and coordinates their planning through the market. What do I mean by this? Essentially, plants with “a lot” of water (maybe it snowed more than everyone expected in their catchment area) bid in to the market cheaply. Plants that don’t have much water (maybe it snowed much less there than everyone expected) bid in at high prices. All the bids are put into a “pot” and selected in order of price (lowest first) until the demand is met. The cheaper the bid the more likely you are to be selected. As demand varies a bit from hour to hour, you can see that cheap bids will be selected almost all of the time, and produce a lot. Mid priced bids will be selected some of the time and produce sometimes. High priced bids will only be selected when demand is really high, and will produce rarely. In this way, plants with lots of water produce a lot, and plants that have little water produce only when absolutely necessary. In this way we ensure we have a flexible power system that can respond to any need, and where we don’t have to cut power to your house in the middle of winter.
So, how do the power companies figure out their market bid prices? Surely low, mid and high are relative, so why are they 200, 500, or 1000 NOK/MWh (which is what they often are) instead of, say, 0, 0.1, and 0.2 NOK/MWh? The answer is that we don’t have enough hydro power in Norway to cover us in all eventualities. In dry years we simply do not produce enough to meet demand, both in peak periods and over the year as a whole. So, we had (and have) two choices:
1. We could build out many many more hydro reservoirs and power generators, so that we could cover our power needs even in the driest of dry years,
2. We could interconnect to other power markets with different types of production (wind, gas, nuclear etc) and buy power from them when we don’t have enough.
Guess which is cheaper (by a huuuge amount) and doesn’t require destroying much of Norway’s nature? It should be no surprise that the answer is not #1.
So, we go for #2 - we build cables. This has an extra advantage that we can sell excess power to our neighbours when we have too much - in wet years for example. Bid prices are then driven by the cost of importing power. Essentially, if I have a power plant with a reservoir, I can calculate a bid price by answering the question: if I were to remove a “unit” of water from my reservoir and throw it away, this would mean I would no longer be able to generate power with that water at some point in the future. Someone else would have to do it - and quite likely a power plant in another country, imported on the interconnector cables. So my expected price of that imported power (“expected” because it’s in the future, I don’t know what it will be) helps me calculate the value of that unit of water. And it is that “water value” I use to bid into the market.
FACT BOX: What if we didn’t have cables, and didn’t build out any new plants? This would mean that in dry periods, we would sometimes run out of power. The water values would then be determined not by the cost of imports, but by the cost of running out of power. What is the cost of blackouts? It is high - interrupted businesses, cold homes, a damaged economy. It’s hard to put a precise number on it, but it is high. The principle for calculating water values though is the same, but instead of, say, a 750 NOK/MWh cost of import driving water values, it would be a 10000 NOK/MWh cost (damage) of cutting load to consumers, businesses and industry. It’s easy to see that this would result in higher power prices than with cables. The only alternative if we “cut the cables” would be to build many more power plants. But who pays for these? We would. There’s no free lunch - or free power.
Each power company uses complex modelling and analysis to calculate these water values and bids. It is a hard task - it takes lots of computing power, and is mathematically very hard. But by splitting it up so that each company can do it for their own assets only, we get a bit of “wisdom of crowds” effect - each might individually be wrong but on average not too bad. And because things are uncertain, we gain a benefit from having many small guesses instead of one big one (which if it is wrong can be catastrophic).
This price is the “right” price to ensure that we export only what we should when we have too much power (e.g. in wet years), and that we import only what we need and as cheaply as possible when we don’t have enough (e.g. in dry years). If I priced too low I would generate too much and end up running out of water. If I priced too high I would generate too little and have to spill water. This is where the third planning problem - uncertainty - comes in. I don’t know what the price of power in my neighbours will be. I don’t know how much water will flow into my - and everyone else - reservoirs. I don’t know how cold it will be and how much electricity will be demanded and when. I can (and do) guess, but I don’t know.
BTW, this uncertainty is what drives quite a bit of the import and export on the cables. As I get new information - measurements of snow in the mountains, a new weather forecast, an updated price of gas and hence a new expected power price in Germany - I will update my “water value” and thus bid price. Since all power producers are doing this, the price tends to fluctuate up and down, and this drives a bit of import here and a bit of export there. For example, in a given year we may export net 5 TWh of electricity, but that may be made up of 15 TWh of export and 10 TWh of import at various stages over the year.
So, the market is essentially a mechanism for coordinating the use of Norway’s hydro resources. But is it the most effective way of coordinating them? Couldn’t we just put a bunch of smart people in a room with some computers and let them do it instead?
Well, we could. But how would they do it? They would have to create the same type of models of the power system that each power company uses today to calculate their bids, to calculate the planning and scheduling of each power plant. But they would have to do this for all plants at once, instead of (as today) for the smaller group of plants owned by an individual power company. This is actually such a challenge that in countries that have done this central coordination (either now or in the past, Norway included), they replicate the market approach by creating an internal market which each asset “bids” into. Why? Because this is a very efficient way mathematically of solving the coordination problem. Indeed, if we deep dive into the maths (don’t worry, I won’t subject you to that…), a marginal-price based market and a plant-by-plant schedule are equivalent. They are the same. You can take one (e.g. a market) and figure out the other (e.g. the production plan) and vice versa. They are equivalent solutions to the same problem.
The key essential difference to these two approaches - a market verses central planning - is that in the latter the market bit is “hidden”. The price for power exists in both - it is a result of the fact that we don’t have infinite electricity, and so we need to somehow plan and coordinate production and consumption. The main difference is that in a central planned system the public isn’t aware of the “price for power”, and politicians don’t use it to bang each other over the head. But, they both imply a price for power - it’s just that the market makes it visible.
True, in the central planning version you don’t need to charge that price to consumers. But neither do you in the market version - see the “Norgespris” for an example of that. This is especially so in countries - like Norway - where the state owns the power production to a very great extent and where taxation is used to recover a good chunk of the market profits from power companies. There are many discussions in how best to return that to consumers to compensate them for their power expenses - e.g. through a price cap, or through a “power dividend” for example - which are beyond the scope of this discussion. But they key point is that no matter what is done to shield consumers from the “price of power”, this price exists, and it is not equal to a few øre per kWh.
Indeed, the market version has some extra benefits over the central planning one. Firstly, if a mistake is made in the central planning approach - e.g. they get it wrong about future inflows - it can be catastrophic. Because they control all power production, any such errors get multiplied out to production plans for all power plants - making it more likely that we do not have enough power (and have to import more expensive power from elsewhere) or we have too much even to export and we spill it. By splitting up the decision making amongst many smaller companies (still state owned), you minimise this risk. If I get it wrong and guess too high, you might get it wrong and guess too low, and the system ends up not too badly off.
In addition, in a centrally planned approach there is a tendency (in practice) to over invest in new production - because this is easier than solving perfectly the planning problem. This costs money, and means the power system is more expensive than it needs to be. This has to be paid for by taxpayers - us. If not in power prices, then through other mechanisms, like grid fees and higher taxes. But it will be paid for - it has to be. Power plants don’t grow on trees. Markets enable a commercial investment approach - which generally means less over investment (there are other challenges with commercial investment, especially in a market environment, but that’s a topic for another time).
So, to conclude this somewhat longer-than-expected essay. The market is simply a pretty effective way of coordinating production (and increasingly consumption) of electricity. So we have power available when and where it is most valuable, and so we don’t run out of the stuff. We could do this coordination in other ways (eg. central planning), but this is harder, more prone to errors, and where individual errors have bigger negative consequences. Cables help us with this market - by giving us access to power when we need it, and a place to sell our power for profit to Norway AS when we have too much, and by helping us to price accurately and thus coordinate as best as possible, and by reducing the number and size of power plants we need to invest in (and any consequent destruction of nature).
There is a challenge when power becomes expensive - but this challenge is there no matter whether we have a market or not. Someone has to pay for it, and ultimately that is us, the taxpayers. Power is expensive for many reasons, including because it is hard to build new generation (e.g. wind in Norway) and because cheaper generation has been closed in many places (e.g. nukes in Germany). There may be very good arguments behind each of these, but they have a singular impact - higher power prices. There ain’t no free lunch in this business. But the market itself is not the reason, it’s just making it very, very obvious. And it enables Norway to get the most out of the power generation it has at lowest possible cost, as a mechanism to coordinate and plan production (and consumption) to keep the lights on and supply power where it generates the most economic benefit for Norway. If you got rid of the market, it would not disappear, just become invisible. And less effective - which would be a worse solution for everyone.
So, getting back to our original question - why doesn’t hydro power cost us basically nothing?
The answer is that the “power price” is a result of us having to coordinate how we produce, consume, import and export power to ensure we don’t run out of it and that we use it when and where is it most needed. Think of it as our best estimate of the “value” of power to Norway AS at any given time. This value, or price, exists irrespective of whether we have a market or not. But the market makes it visible. The price would implicitly exist anyway no matter how we planned and operated our power system.
Having cables, importing and exporting, or not, doesn’t remove this underlying price - it just changes what this price would be. Cutting cables just means we would either run out more often (in which case the price of power would on average be even higher than today) or we would have to build out large amounts of new production (with all the cost and environmental destruction that would entail). How we chose to expose consumers to this price is a separate question. We could subsidies and have fixed prices (e.g. Norgespris), or we could expose consumers to these prices. These are political decisions (with economic implications). But in any case, the value or “price” of power, even in hydro systems like Norway, is not zero or near to it.
Comments