The demand for electric energy steadily increases and the power companies predict a substantial deficit in our energy production and propose new nuclear or fossil fuelled power plants. This increase in the demand is mainly due to the increased use of electric energy for heating applications, creating the need for additional energy imports particularly in winter months. Much of this additional imported energy is produced by fossil fuelled power plants.
The energy demand for heating is large, in 2005 our country burned 6238 millions of litres of oil and in energy equivalents over half of this amount in gas . In comparison: the average amount of water flowing in the Rhine in Basel is about 1 million litres per second, the amount of fossil fuel burnt for heating every year corresponds to over two hours of the water flowing through the Rhine!
Energy for heating is subject to the peak demand in winter. To illustrate the situation I measured the weekly demand of energy for heating and hot water generation in my house in Saint-Aubin (NE). Applying the monthly percentages of the annual total as measured in my house to Switzerland’s total energy consumption for heating covered by fossil fuels of 78’731 GWh  we get the average monthly energy used for heating our country covered by fossil fuels as shown in graph 1.
The graph shows the enormous amount of energy used for heating and sanitary warm water generation: in January over 15 TWh corresponding to over 21 GW mean power. The situation is in reality worse because the figures in the winter months neither consider the annual fluctuations nor the fluctuations within the month. During the coldest days the power requirement is therefore higher than the above figures indicate. The demand may be better appreciated in comparison with the production of our nuclear power plants: the same year (2005) they produced a total of 22 TWh .
The figures show that the substitution of fossil fuels by electric energy will require enormous investments in generating and transportation capacity. Fortunately, resistive heating is progressively abandoned in favour of heat pumps. According to ref. [3, 4] the yearly performance factor (COP, ratio between thermal energy output and electric energy consumption) of these machines is around 3. Assuming a COP of 3, the power demand in January reduces to an average of 7 GW, which is still more than the annual average power of our total electric energy production (in 2005: 6.3 GW).
Our hydraulic potential being essentially exploited already, additional power will require nuclear or fossil fuelled power plants. Both are subject to the thermodynamic efficiency limitations of the Carnot cycle which lead to the fact that 2/3 of the primary heat energy is wasted and dumped in to our environment in the cooling towers of these plants. As an illustration: the Leibstadt nuclear power plant has an electric output of 1.2 GW, roughly twice this amount of power is going in the cooling tower as heat energy with a temperature corresponding to our heating requirements. The waste heat from Leibstadt dumped in the environment in the cooling tower could thus heat over 10% of Switzerland’s fossil fuel heated homes even in January. Why is this energy saving potential not realized? The answer is simple: heat cannot be transported over larger distances in an economically defendable way.
The conclusion from this situation is obvious: if heat from thermal power plants cannot be transported to the consumer one has to install (small scale) power plants at the location where the heat is used and transport the generated electricity. This however requires a paradigm change in our habits of electricity production and consumption: the end user becomes at the same time producer and the electricity grid is not used in one way only (from large production units to the consumer) but will serve on a local level as a distribution system used in both directions.
To obtain experimental data of a real application, I replaced in 2004 the classic oil boiler of the heating system in my house by a small cogeneration machine. Several machines for this purpose are commercially available running on fuel oil or on gas and the system is quite common in other countries (e.g. Germany). The principle of these machines is simple and relies on well known technologies. Fig.1 shows a schematic representation. The fuel (oil or gas) is burned in an internal combustion engine which is coupled to a generator. The waste heat from the engine (in the cooling water and in the exhaust gas) is collected and used to heat the house. The electric energy is in this case a (very valuable) by-product used directly in my house, the surplus is fed in the electricity grid. Figure 2 shows the machine (with removed cover panels).
I opted for a fuel oil burning system (since I had an oil infrastructure from my old heating system) of 5.3 kW electric power. The cogeneration machine is my only source of energy for heating and sanitary warm water and the performance of the system may be seen in graph 2 for 2008/2009.
The graph shows, that
1. The cogeneration produces substantially more electric energy than my house uses (annual average).
2. The supply to the grid is to 120% in the winter period; in January 2/3 of the electricity production is supplied to the grid, in summer I am consuming from the grid.
3. The total production of electric energy would allow to heat a similar house with a heat pump having a COP of 3, in January my supply to the grid would still allow heating a house with a heat pump having a COP of less than 4.
Ecologic and economic aspects
Compared to an ideal boiler type heating, a cogeneration system theoretically burns a surplus of fuel corresponding to the output in electric energy. This is however in practice by far not the case. Compared to my boiler type heating which used 2588 L of fuel oil (20 year average) I burned in 07/08 2600L and in 08/09 2800 L of oil, which is a far smaller increase (due to the losses inherent to a boiler). Even though my data show better performance, I assume an additional fuel consumption of 20% compared to a boiler type heating. This additional fuel however produces 30% of electric energy. Replacing boiler type heating by minicogeneration systems thus allows producing electric power with an overall conversion efficiency of 150% which clearly is the most efficient way to produce electric energy from fossil fuels. To illustrate its potential for Switzerland: replacing half of the boiler type heating systems in our country by cogeneration machines would allow producing the power to heat the other half of the houses with heat pumps. This would lead to an overall reduction in the fossil fuel consumption and CO2 emission of 40% (half the total number of systems, but every system burning 20% more than a conventional boiler), which corresponds to the most ambitious climate goals actually discussed.
The investment in a mini-cogeneration machine is of the order of 30 000 CHF, compared to the investment in the Leibstadt nuclear power plant one finds, that the investment per kWel is of the same order. Mini-cogeneration machines are almost unknown (in Switzerland). The technology is mature and suitable for mass production and therefore has the potential for cost reduction. Decentralized power production does not need additional transportation capacity thus making the cogeneration also an interesting option considering the investment needed to cover the additional power needed in winter.
 Überblick über den Energieverbrauch der Schweiz im Jahre 2005, BFE, Juni 2006, Bundespublikation Nr. 805.006.06
 According to  assuming the whole consumption of fuel oil and half of the gas consumption is used for space heating and hot water generation.
 Road map erneuerbare Energien Schweiz, Schweizerische Akademie der Technischen Wissenschaften, SATW Publikation Nr. 39
 Markus Erb et al. Feldanalyse von Wärmepumpenanlagen FAWA 1996 - 2003, ENET Best. Nr. 240016
 Picture courtesy of Senertec GmbH, Schweinfurt
[Released: June 2010]