The spillway at the Robert-Bourassa hydroelectric power station in Québec. (Facebook | Eeyou Istchee Baie-James - touristic region), CC BYHow much will a kilowatt-hour (kWh) of electricity cost in 2040? The short answer is: a lot. How much would it cost to preserve the energy service of that kWh without actually generating it? Less — much less, in fact.You can heat a poorly insulated house to achieve a certain level of comfort. Or you can insulate the house and achieve the same level of comfort while using less energy. Given two options that offer equivalent comfort, most people would normally opt for the cheaper one.What goes for a house also applies to a region — and to technologies other than heating and insulation. In all cases, if the cost of generating a kWh is higher than the investment required to obtain the same service with less power — let’s call this the “cost of efficiency” — we will choose efficiency. The difference between these two costs is called the “value of saved power.” The higher it is, the more valuable efficiency becomes, compared to generation:Cost of generation - cost of efficiency = value of saved powerTogether with researchers Florian Mitjana and Pierre-Olivier Pineau from HEC Montréal, I assessed the value of saved power for the electrical system of the northeastern part of North America: the New England states, New York state, Ontario, Québec and the Atlantic provinces. The Town Hall of La Pêche, northwest of Ottawa, built entirely of wood, is a cutting-edge example of energy-efficient, low-carbon architecture. (Facebook | Latéral), CC BY Power and the energy transitionClimate change is largely caused by fossil fuels. So most countries are trying to reduce their use, albeit with uneven levels of commitment.This move away from fossil fuels takes largely place in favour of electrical power, meaning demand for such power will grow much faster in the coming years than it has in recent decades.Moving away from fossil fuels also means new electricity must be generated without resorting to coal- and gas-fired power stations. Electrical power generation must therefore be boosted while replacing a significant number of power plants. This planning headache is known as the “energy transition.”Generation cost: A future marginal costAssessing the cost of generating a kWh, the first part of our equation, isn’t easy. That’s because the current average generation cost isn’t an adequate measure:Current costs are misleading because the energy transition has a huge impact on how electricity is generated. Instead, we must anticipate future costs, accounting for the transition discussed above.Average costs aren’t very helpful here either. The kWhs that matter for this decision — those we can choose to produce or avoid — tend to be the most expensive. Why? Because not all electricity costs the same to generate. We use the cheapest sources first, like prime wind or hydro sites, and turn to more costly options only when needed. The price of these “last” units of electricity is known as the marginal cost.That means that the future marginal cost is the best measure of generation costs. Calculating this involves foreseeing how the transition will unfold. What kind of power grid will be built over the next, say, 15 years? How many wind turbines will have been built? Where? In Québec, Nova Scotia? Or offshore? How many gas-fired power plants and solar panels will there be? Where will high-voltage transmission lines be built? Will batteries be used to store wind and solar energy, and at what scale?ChallengesThere are tools available to calculate the future marginal costs, including mathematical models, optimization algorithms and databases. But they’re just tools, not magic wands. Many difficult decisions must be made in the process. Two major challenges: What will the power demand be in 15 years?This question strongly influences our results, but it’s difficult to answer. We can (temporarily) get around the question by simply calculating a future marginal cost for different demands.How can we anticipate the transition? How can we know what electrical power system will result from the decisions of thousands of political and private entities? A reasonable solution is to assume that stakeholders will collaborate and opt for the lowest-cost decisions. That requires taking into account several constraints: the wind does not always blow when we want it to, the best rivers have already been harnessed, the use of gas-fired power plants must be restrained to reduce emissions, and so on.The future marginal costs we obtained are shown in the following figure. The horizontal axis shows different future demands for 2040. The value in brackets is the percentage of the 2025 demand (from 105 per cent to 200 per cent). The blue bars are 100 terawatt hours (TWh) wide. (The power system under study is so large that kWhs become terawatt hours.) The height of a blue bar indicates the marginal cost of generating an extra 100 TWh, in U.S. dollars per kWh.For example, if total demand increases from 866 TWh to 966 TWh, the cost for this 100 TWh increase is calculated at 42 American cents per kWh. The striking result is that the marginal cost increases very rapidly when the 2040 demand reaches 130 per cent or 140 per cent of the 2025 demand. If demand doubles, the marginal costs skyrocket to values above one American dollar per kWh.The cost of efficiencyLet’s now turn to the cost of efficiency, the second term in our equation. It’s necessary to determine the cost of reducing the amount of power of an energy service; it could be, for example, the cost of caulking the doors and windows of a house. À lire aussi : La menace méconnue de la transition énergétique This problem is easier to solve because the cost of efficiency isn’t driven by the energy transition or by the level of power demand. So we can rely on the best current studies to make our calculation.According to our review of the literature, a fairly reliable cost of efficiency is 10 cents per kWh, regardless of future power demand.The value of saved powerWe are just one subtraction away from determining the value of efficiency. Simply take the marginal costs from Figure 1 and subtract from them the cost of efficiency — 10 cents. The results are shown in Figure 2.Figure 2 reads in the same way as Figure 1. The cost of efficiency has been added, shown as the horizontal line at almost 10 cents per kWh. The value of saved power is given by the height of the red lines. For example, if we want to add 100 TWh of electricity to a system of 866 TWh, it costs 42 cents per kWh, whereas if we add 100 TWh of energy service through efficiency measures, it costs 10 cents per kWh. The value of saved power is the difference between the two — so 32 cents per kWh.In other words, choosing to meet 100 TWh of energy demand by generating electricity costs 32 cents per kWh more than it would through efficiency. That means consumers are paying an extra $32 billion, annually, for the same energy service. Worse still, if demand turns out to be higher, the value of saved power quickly reaches one dollar per kWh. Lessons to learn, time to actIt’s interesting to revisit the challenges discussed above and draw conclusions.Forecasting electricity demand in 2040 is key. We estimate that a 50 per cent increase between 2025 and 2040 is plausible, with the possibility of slightly lower growth, but also the clear possibility of an increase well above 50 per cent. That leads to our main scientific conclusion: We are heading towards a future where efficiency will have a high value. For policymakers, that means it’s critical to invest now in a more efficient economy.In terms of anticipating how the energy transition will proceed, we made the bold assumption that the five regions in our study — comprising seven U.S. states and six Canadian provinces — would co-operate optimally. This assumption may be unrealistic, but it doesn’t change our conclusions — on the contrary.A basic principle of optimization implies that when co-operation is imperfect, marginal costs rise. This is not the case for efficiency costs, which don’t depend on co-operation and remain at 10 cents per kWh regardless of regional co-ordination. As a result, the value of saved power would be even greater than we estimate — further reinforcing the conclusion above.The energy transition will not happen without energy efficiency. At least not at a reasonable price.This research was conducted in part thanks to financial support from IVADO and the Canada First Research Excellence Fund.