Author: Aleksandar Macura
How free is the choice of heating method if citizens cannot see what they are actually choosing?
More than ten years ago, in the debate on competition in the heat market led by Fortum and Oxera, a fairly simple test for fair competition between different heating methods was formulated.
The consumer should be able to choose the heating method. He should receive transparent and objective information about the alternatives. Connecting and disconnecting from the district heating should be voluntary and simple. And taxes and public subsidies should not arbitrarily favor one technology over another.
When we think like that, price comparison does not mean just comparing the price of gas, electricity or wood.
It includes fixed costs, taxes, investment in equipment and installations, costs during the lifetime of the equipment, and environmental costs.
We need better information, real choice and fair competition between heating methods. If we tell citizens to choose, we must enable them to really see what they are choosing between.
Let’s take a look at the information we offer citizens about heating costs in Serbia today and ask: have we enabled citizens to see what they are actually choosing between? Can citizens choose?
Which car is better than Moskvich?
What does the state (not) tell us when it (doesn’t) tell us how much heating costs?
A private choice of heating method has public consequences. That’s why the information on the basis of which citizens choose how to heat themselves is neither harmless nor neutral.
Two years ago we were already knocking on the barometer.
The occasion was the regular annual announcement of the Energy Agency of the Republic of Serbia on energy costs for heating residential space. It reminded us of the barometer from Three Men in a Boat which, while it was raining outside and the river was overflowing, persistently showed fine and dry weather.
Our energy barometer has shown mostly good gas weather for years. Natural gas was the cheapest or almost the cheapest way of heating even while things were happening in the European markets that can hardly be described as good gas weather. Then we noticed something else: the barometer didn’t show all the weather conditions. There was no district heating in the AERS comparison. There were no heat pumps either.
Two years later, the pointer still works, and now we’re going to take a look at what’s actually drawn on the counter.
The annual AERS announcement seems like a completely neutral piece of information: this costs this much, that costs that much, so you, the public, take your pick. But the choice of heating method is not only a private matter, nor is it done in a self-regulated market.
The furnace, boiler or heat pump we buy today determines what we will spend in the next fifteen or twenty years. Our choice affects the quality of the air we all breathe, the import of energy we all pay for, the load on the electric power system, emissions, forests, local infrastructure and public expenditures with which the state reacts tomorrow when some energy becomes too expensive.
An apartment of 60 square meters in which the whole of Serbia lives
AERS bases its calculation on a hypothetical household.
The apartment has 60 square meters. The whole is heated 16 hours a day, 180 days a year, at 20 degrees. For heating, it needs 9,000 kWh per year, or 150 kWh per square meter. Then, depending on the assumed efficiency of the device, it is calculated how much wood, gas, pellets, coal or electricity is needed.
For the heating season 2026/2027, we thus receive, among other things, about 60,000 dinars for cheaper wood, 59,500 for natural gas, 113,900 for pellets, 102,700 for a TA stove that is charged exclusively at night, and 250,600 dinars for direct electric heating.
Everything looks neat. Numbers have the last hundred dinars. Efficiencies have percentages. Consumption has kilowatt-hours.
Computational accuracy and image accuracy are not, however, the same thing. You can accurately calculate the consequence of an assumption that describes the real world very poorly. Let’s start with 150 kWh per square meter.
The report of the Business Association “Toplane Serbia” for the year 2024 offers an interesting check on this assumption. In dozens of district heating systems, the specific consumption ranges from approximately 74 to more than 270 kWh per square meter per year. But when the systems are not treated as equally large, but the consumption is weighted by the actual heated surface, an even more interesting result is obtained: around 112 kWh/m² per year. This does not mean that we have found a new “real average consumption of Serbia”. District heating users are not a representative sample of the entire housing stock, and the delivered heat in the district heating system is not necessarily methodologically identical to AERS’ estimated heat demand. But when for almost 48 million square meters of space we have a real annual delivery that gives about 112 kWh/m², then 150 kWh/m² can no longer be a self-evident assumption. The regulator owes the consumer at least an answer as to where that number comes from, which part of the housing stock it refers to and how much the result changes for buildings that consume 80, 110, 150 or 200 kWh/m².
Heating that exists, but is not there
An even bigger problem arises when we look at what is in that table in the first place.
There is no district heating in it. The document that compares the costs of energy for heating to citizens does not show the costs of one of the most important methods of heating in Serbian cities.
OK, there is no single price for district heating in Serbia. Prices are determined by local governments, tariff systems differ, sometimes you pay per square meter, somewhere according to consumption, there are fixed and variable parts.
But there is no price for wood in Serbia, no price for coal, no price for pellets. AERS already shows cheaper and more expensive wood.
AERS could show the cost in several large district heating systems, median, range or representative examples. He could at least tell the citizen: if you are connected to the heating plant, this is the way you can compare your costs with other options.
Instead, the heating plant simply disappears from the counter.
And then the most efficient way of electric heating disappears
AERS is much more careful when showing electricity, so it offers us consumption displays for a TA storage heater that only charges at night, then for a TA storage heater that is topped up during the day, and for direct electric heaters and boilers. In the latter case, the calculation for our imagined apartment amounts to more than a quarter of a million dinars per year.
However, there is no heat pump. There is not even its most widespread form, which hundreds of thousands of citizens know under a much less exciting name – inverter air conditioner. It’s not there… It’s just not there. It is an unusual non-disclosure choice made by the regulator.
If you show a citizen how much it costs when you convert one kWh of electricity in a heater into approximately one kWh of heat, and you don’t show a device that uses the same kWh of electricity to transfer several units of heat from the outside air, water or earth, you haven’t shown him “how much electric heating costs”. You showed him how much one of the least efficient ways of using electricity for heating costs.
Heat pumps are not some exotic innovation that surprised the state administration. The Republic of Serbia subsidizes their purchase through energy rehabilitation programs. State projects keep data on the number of installed devices and their capacity.
The heat pump is real enough that the state gives the citizen money to buy it, but it is not yet real enough to appear in the regulator’s annual information on heating costs. It is a bad form of public information.
The stove is as efficient as we say it is
With wood, the problem is different.
AERS uses a device with an efficiency of 65% for the cheaper wood, and 55% for the more expensive variant.
It looks neat on paper.
It often looks different in the living room.
The RES Foundation has been dealing with the efficiency of individual solid fuel appliances for much longer than this text has existed. In our Smarter Stoves regional survey, we reported informed estimates that the actual efficiency of a large portion of existing appliances is around 30-40%. At the same time, 65% is the minimum seasonal efficiency that is linked to certain new eco-design devices. The actual efficiency is affected not only by the construction of the device, but also by the humidity of the wood, the chimney and the way the user fires.
The difference is not small. If the device for the same amount of useful heat operates at 35% instead of 65% efficiency, it needs approximately 86% more energy from the fuel. Efficiency can dramatically change the annual bill, the order of technologies in the table, and then perhaps a household’s decision on what to buy.
The extra wood to burn is not just an extra private cost. It also means greater consumption of biomass and, in the case of old devices and poor combustion conditions, greater emissions of pollutants.
A seemingly private decision produces public consequences.
AERS, honestly, also says what it doesn’t count
AERS explicitly states that the amounts shown include only the procurement of energy or fuel. They do not include the purchase of a furnace or boiler, the installation of central heating or periodic maintenance. It is an important fence that is also a description of the problem.
When a household chooses how to heat, it doesn’t just choose a cubic meter of gas instead of a kilowatt-hour of electricity. Selects equipment. Connection fee. Pays for chimney, service, installation, fuel space. He changes some devices more often, some less often. Somewhere he pays a large initial cost and a small energy bill, somewhere the other way around.
Therefore, the cost of fuel is not the same as the cost of heating. And neither of those two is yet the same as the cost of heating to society.
The state does not stand on the sidelines while the citizen chooses, but it touches the price tags
We could consider everything previously said as a technical criticism of a nice annual table. Add a heat pump, find a better average, introduce several scenarios, put heating plants and the job is done.
However, the heating method is not a consumer decision like choosing a brand of detergent.
When a household installs a gas boiler, a wood-burning stove or a heat pump, it enters into a technological choice from which it will not easily return for decades to come. During that time, the state determines the prices of network energy sources, taxes, excise duties, connection conditions, building regulations, subsidies for the purchase of equipment, subsidies to energy companies, social protection, device standards and much more.
The state, therefore, already very strongly regulates the terrain on which free private choice supposedly takes place.
The information on heating costs it provides is just another instrument for guiding private choices.
That’s why it doesn’t matter what the state institution puts in the table, what it excludes from it and what assumptions it presents as average.
At the beginning, we said what is needed for fair and real competition between heating methods: the consumer must have freedom of choice; must be informed and engaged, with transparent and objective information about the prices of alternatives; connection and disconnection from district heating should be voluntary and simple; and taxation and investment support should not arbitrarily favor certain alternatives. The price comparison should include not only energy but also fixed costs, taxes, environmental and capital costs over time.
If we tell citizens to choose, we must enable them to really see what they are choosing between.
What would a fairer barometer look like?
We are not asking AERS to know how much every house in Serbia will be heated next winter.
Instead of an imaginary apartment, there could be a simple public calculator with several typical levels of required heat or, even better, the possibility for the user to enter his own consumption.
Instead of one device efficiency, there could be a range: old furnace, new furnace, modern boiler.
In addition to direct electric heating there would have to be a heat pump, with a clearly stated assumption of the seasonal efficiency factor.
District heating could be shown across a range or representative local systems.
The annual cost of energy, investment, fixed costs and maintenance would be shown separately.
And with all that, there could be information about whether the state subsidizes a certain investment.
We could get a much fairer representation of reality.