What sort of financial support mechanism would best promote the introduction and use of micro-CHP units as a method of cutting carbon emissions? A.D. Hawkes takes a look at how poorly-designed support could lead to gratuitous use of the technology – with no carbon benefits at all.

Micro cogeneration (or micro-CHP) is a class of small-scale electricity and heat generation technologies designed to replace existing residential space and water heating systems. A number of studies have shown that these systems have the potential to reduce greenhouse gas emissions associated with residential energy use, possibly at relatively low cost. This, combined with their significant market potential, has generated a great deal of commercial interest. Arguably the most successful commercial systems to date are based on internal combustion engine prime movers in Japan and continental Europe, but there has also been a great deal of recent commercial interest in fuel cell based systems (predominantly polymer electrolyte membrane (PEMFC) and solid oxide(SOFC) varieties), Stirling and Rankine cycle units.

Systems typically consist of a prime mover, a supplementary thermal system such as a condensing boiler (to meet heat demands at peak times or when it is not economic for the prime mover to do so), related thermal management hardware, control systems, fuel reformers, gas-air delivery systems, and afterburners for the case of fuel cells.

Policy to support the introduction of micro cogeneration and other residential microgeneration products such as solar PV, biomass boilers, heat pumps and micro wind turbines is rapidly emerging. For example, in the UK, there is a consolidation of residential energy efficiency policy and microgeneration policy in progress, with the primary instrument requiring energy suppliers (i.e. the tertiary retail element of the liberalized energy market) to achieve uptake of ‘qualifying actions’ (energy efficiency and microgeneration) in their residential customer base in order to meet an aggregate CO2 emissions reduction target (CERT). This emulates a greenhouse gas cap-and-trade system, although does not necessarily provide an absolute emissions reduction because only a reduction quantity is specified, rather than a cap.

While the CERT is a reasonably economically rational instrument in that it allows suppliers and owner-occupiers to choose a solution that fits their budget and situation, it is widely seen as insufficient to boost the uptake of microgeneration. Consequently, instruments such as grant support and VAT (goods and services tax) reduction are also available.

Of particular relevance to this article are topical discussions regarding the application of a feed-in tariff to support micro cogeneration uptake. Indeed many suppliers in the UK offer buyback rates for microgenerators that are very generous (but limited to a niche market at present), creating a situation similar to what would exist with a mandated feed-in tariff in a mass market.

Perhaps the primary example of this approach is the German one, where around €0.4/kWh feed-in tariff is applied to solar PV and some other renewable systems, resulting in one of the most impressive PV penetrations in Europe. This article considers the incentive provided by a feed-in tariff for micro cogeneration in terms of economic emissions reduction, highlighting challenges in such an approach.


Firstly, consider the basic economics of micro cogeneration. Take for example the case where a 1 kWe system with 30% electrical efficiency (i.e. a near to medium-term fuel cell-based system) and 90% overall (heat plus power) efficiency meeting various kinds of loads. Here we assume energy prices of 3 pence/kWh for gas and 10 pence/kWh for electricity. Electricity buyback is offered at a rate of 8 pence/kWh (which, in reality, corresponds to the wholesale price of electricity plus a feed-in tariff of about 4 pence/kWh). For the UK the grid emissions rate is assumed to be 0.43 kg CO2/kWh, and consumption of gas produces 0.19 kg CO2/kWh. Table 1 shows the basic economics for a variety of load combinations.

Click here to enlarge image

The first two demand scenarios indicate that this micro cogenerator certainly does have the potential to provide emissions reduction and operational cost savings. Depending on how long the generator operates in each mode, and the system’s capital cost, positive overall economics is a possibility. However, the most interesting result from a policy point of view is the final line in Table 1. If a micro cogenerator can dump heat from the system, with these efficiency values and price combination, it is irrelevant from an economic point of view whether you meet demand via the micro cogenerator (and dump the heat) or you meet demand using grid electricity.

Most importantly, the corresponding environmental result is significantly biased towards using grid electricity, which produces almost 50% less CO2. Obviously this is a disastrous result for the micro cogenerator, undermining one of the main drivers for government support of the technology.

A key question in relation to this issue is: does micro cogeneration have the technical ability to dump heat? Certainly this is technically possible, but in practice a facility to pipe heat out of a dwelling in order to dump it is likely to be expensive (and clearly environmentally questionable, implying minimum standards regulation may be formulated to prevent it).

However, while deliberate technical heat dump is unlikely, an owner-occupier has several options to achieve the same result. The simplest of these is to open a window, or otherwise increase the rate of turnover of air inside the dwelling. Alternatively one could turn up the thermostat. Furthermore several longer-term options to increase heat demand unnecessarily are available. These relate to deferred or avoided upgrade of insulation or glazing, removal of draught stripping or changes to control of the overall HVAC system.

Undoubtedly, increase of heat demand is possible, and it is not difficult to interpret such actions as accomplishing heat dump. Questions remain regarding whether or not a typical owner-occupier would equate increased demand with equal (or lower) cost, but perhaps the fact that increased ventilation is usually associated with increased thermal comfort would encourage such action.


The case considered up to this point has been very simplistic regarding depiction of the micro cogeneration system, and its operation. Now a more sophisticated tool will be applied to consider this problem. A technically-rich characterization of a variety of micro cogeneration systems is considered, along with an analysis of how it would perform in a real dwelling over the course of an entire year.

The tool used to examine this situation is a mixed integer linear programming optimization framework known as CODEGen (Cost optimization of decentralized energy generation), developed at Imperial College London. This tool minimizes the annual cost of meeting a given residential energy demand profile by considering the operational schedule of the generation system (i.e. a classic unit commitment problem).

In this case the micro cogenerator is defined via a steady state efficiency profile (including part-load performance), and constraints such as ramp rate limits, minimum up time and down time, start and stop costs to emulate system dynamic behaviour.

Once the optimum operating schedule of the system is known, the cost of meeting energy demand using the micro cogenerator can be compared to the cost of using the reference system, resulting in the ability to estimate the value of the cogenerator relative to the reference system (i.e. the maximum amount of money a rational investor would pay for the cogenerator over and above what they would pay for the competing boiler-only system).

The annual CO2 savings can also be calculated in comparison to the reference system.

Figure 1 shows the results of applying this modelling approach to consider four key micro-CHP technologies: internal combustion engines (ICE); polymer electrolyte membrane fuel cells (PEMFC); solid oxide fuel cells (SOFC) and Stirling engines. Each technology has specific characteristics based on evidence from field trials and theoretical modelling, but the key difference in terms of this study is their electrical efficiencies.

Figure 1. Relative value and annual CO2 reduction for four key micro-CHP technologies for the cases of constraint and no constraint on dumping heat
Click here to enlarge image

The peak load electrical efficiency is 10% for the Stirling engine, 20% for the ICE, 30% for the PEMFC, and 40% for the SOFC. This value can vary between manufacturers of specific systems and is therefore only a guide and should not be construed as the only possible efficiency for a system of that type.

Figure 1 shows the economic value and CO2 reduction associated with each technology across a range of buyback prices, for cases of constrained and unconstrained heat dump.


Firstly, the top two sub-plots in Figure 1 are considered. These relate to the case were heat dump is constrained. All systems, except the Stirling engine, exhibit strong dependence of value on buyback price. The higher the electrical efficiency, the greater this relationship becomes. This is because the low electrical efficiency systems produce relatively more heat, and because heat cannot be dumped this has the knock-on impact of requiring these systems to modulate of shut down over longer periods of time and thereby export less electricity to the grid.

This generally reflects the basic economic driver for micro-CHP; more electricity production correlates with better economics, but thermal constraints limit the ability of the system to achieve this electricity production.

In terms of CO2 reduction (top right sub-plot in Figure 1), increasing buyback price also drives CO2 reduction with great effect up to 6 pence/kWh, and then further increase in buyback prices does not promote improved performance. This translates to the point where systems have reached a thermal constraint and thus cannot generate more electricity/heat and provide more CO2 reductions.

Now consider the bottom two sub-plots in Figure 1, relating to the case where heat dump is possible. There is an even stronger dependence of value of the systems on buyback price. Likewise to the discussion above this relates to the ability of systems to produce electricity, and this system is less thermally constrained and can therefore operate in a more flexible manner. Clearly in both cases (heat dump constrained and not constrained) buyback price can be a very effective value driver, and could encourage uptake substantially.

However, the CO2 result for the case where heat dump is not constrained is somewhat different. Particularly consider the case of the PEMFC system; increasing the buyback price has incentivized decreasing the annual CO2 reduction. Therefore if a feed-in tariff were publically funded (or effectively socialized through a supplier-regulatory approach), additional public money has been spent in this case for an arguably worse result. Conversely, the CO2 result for the SOFC system has improved up to buyback price of 8 pence/kWh, and is overall better than the result for the case when heat dump is constrained. These conflicting results suggest that interactions between a micro cogenerator’s CO2 savings and the ability to dump heat are largely dependent on its electrical efficiency, although external factors such as energy prices, buyback rates, and embodied CO2 rates for grid electricity also bear upon outcomes.


To consider this phenomenon further, a surface plot of buyback price, electrical efficiency, and system value is shown in Figure 2, for the case where heat dump is possible. The corresponding annual CO2 reduction surface is presented in Figure 3. This broadly confirms the information in Figure 1, but gives more detail as to the efficiency range and buyback prices under which a sub-optimal emissions result can arise, and therefore can be used to comment more broadly on policy incentives.

Figure 2: Maximum capital cost differential between micro-CHP and boiler-only system (£) versus electricity buyback price (pence/kWh) and net system electrical efficiency
Click here to enlarge image

Firstly, Figure 2 shows that both buyback price and electrical efficiency are key drivers of system value. With regard to the CO2 result in Figure 3, for any electrical efficiency value there is little CO2-related justification for a buyback price above 6 pence/kWh as there is no substantial additional CO2 savings above this value.

Figure 3: Estimated annual CO2 emissions reduction (kg CO2/year) versus buyback price (pence/kWh) and system net electrical efficiency
Click here to enlarge image

The second point in relation to Figure 2 is that for the electrical efficiency range of 25% to 40%, there is a possibility that a high buyback price would result more CO2 being produced than at a lower buyback price. This effect is predominant for the very high buyback prices, and electrical efficiency close to 25%. The explanation for this result relates to the situation where the system can profit from dumping heat, but cannot compete with grid electricity in terms embodied CO2 per kWh of electricity.

Higher electrical efficiency systems can also profit from dumping heat, but when electrical efficiency reaches above roughly 40%, the embodied emissions for micro cogenerated electricity is lower than that of grid electricity and thus better CO2 performance is assured.


From a public policy point of view, the situation where additional public (or otherwise distributed across society) money is spent in order to achieve the same or less CO2 reduction could be seen as inefficient. Likewise, it is also important to note that little or no production-based reward (i.e. in this case, a buyback price of zero to 2 pence) also leads to a sub-optimal result in that substantially more CO2 reductions could have been achieved if the buyback price were slightly higher.

It seems that the optimal buyback price is between 4 and 6 pence/kWh, but any deviation from this can lead to a poor outcome in comparison to what could be achieved.

This creates a challenge to policy makers. Perhaps regulation of buyback price is a reasonable strategy, although clearly at odds with free market principles and almost certainly overly bureaucratic. Indeed the ‘best’ buyback price is likely to vary among micro cogeneration technologies, and is a function of energy prices and embodied grid CO2 rates, creating a complex relationship that would be onerous to monitor and maintain.

It is interesting to note that the current wholesale electricity price (which is arguably close to the fair electricity buyback price for micro cogeneration) is of the order of 4 pence/kWh, indicating that the free market may choose the best environmental solution at the moment. However, this depends heavily on the fuel mix, cost of grid electricity, and value of the microgeneration proposition to the supplier, with no guarantees that the situation will not change.

Overall, the results of this analysis indicate challenges regarding effective production-based incentives for micro cogeneration. Any production-based incentive (whether it be for electricity generated, electricity exported, or heat generated) can lead to the described situation because it can incentivize gratuitous operation of the device. Nevertheless, a production-based incentive is also clearly a strong driver for value of micro cogeneration, and a certain level of such an incentive is certainly required to achieve the best greenhouse gas reduction.

It should also be noted that low cost CO2 reductions are not the only driving factor for residential sector energy policy. For example, energy security and fuel poverty are key concerns, as is support for domestic business. Certain adverse (or identical) CO2 results may be warranted if another policy aim is achieved, or alternatively the aggregate effect of an incentive could be positive in terms of CO2 even though some cases are negative.

The issue investigated in this article is simply an important further consideration that should be taken into account in order to formulate balanced and effective policy instruments.

Adam Hawkes is a Research Fellow at Imperial College, London, UK.
e-mail: a.hawkes@imperial.ac.uk