The four systems have been simulated for a one year period and the results were summed up to monthly and yearly values. The total flue gas losses are also calculated during off periods of the burner and thus contain the leakage losses.

The simulation results from figure 3 show that the flue gas losses are in a range between 9% and 16% of primary energy for all systems. For system 2 and 4 the highest flue gas losses are observed, also when considering the absolute values in figure 4. When looking at the leakage losses the influence of the hot water volume in the boiler of system 4 and the combistore of systems 3 becomes visible. System 1 has almost no leakage losses and system 2 can be found somewhere in between.

Fig. 3. Annual flue and leakage gas losses in proportion to the pellet consumption (left). Annual store heat losses in proportion to the energy input of the store and annual burner/boiler heat losses in proportion to the pellet consumption (right). The absolute monthly heat losses for each system can be seen in figure 4. Systems 3 and 4 have the highest total heat losses, but also the largest store volume and pellet consumption. For this reason it is better to compare the respective losses in proportion to the stored energy and the fuel consumption. For example the store in system 3 has the highest absolute heat losses, almost three times more than the other systems, but in proportion to the stored energy system 1 and 2 are much worse. The heat losses are mainly due to the rather poor insulation of the store near the burner and the exhaust gas outlet. System 4 has the lowest relative store losses but the pellet boiler causes high heat losses. The absolute monthly values in figure 4 show that the flue gas losses dominate the heat losses, but the store and boiler/burner heat losses play a major role.

System 3 Qpei=18850 kWh □Flue gas losses □Leakage losses ■Burner heat losses Store losses — -■ І : :

1200

300

400

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

000

800

600

400

200

1200

200

600

1000

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Fig. 4. Monthly heat losses of four combined solar and pellet heating systems simulated for house with annual space heat demand of 87 kWh/m2 located in Stockholm. The store, boiler and burner heat losses do not contribute to the space heating.

_ x

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

1200

1000

800

600

400

200

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

1200

000

800

600

400

200

From figure 4 it can be also seen that the boiler and burner of system 3 and 4 respectively operate also in the summer months when only domestic hot water is required that can be covered by the solar system and an electrical heater as a backup. Consequently the question was how much energy can be saved by a seasonal operation of the pellet heating units. Moreover it was interesting to investigate to what extent the heat losses contribute to the space heating when the store and boiler are placed in the heated area. For this reason two more variants of the systems have been simulated: variant 2 where the system is placed in the heated area; and variant 3, which is the same as variant 2 but with a seasonal operation of the pellet heating units. This means the burner is turned off from the middle of May until the beginning of September.

In figure 5 the total auxiliary energy demand and the effective store heat losses are illustrated for the simulation variants, where variant 1 is the previous simulation variant with the heating system located outside the heated area of the building. The total auxiliary energy consumption consists of the used pellet fuel energy (lower heating value) and the auxiliary electricity. It can be seen that system 1 is the most energy efficient system for variant 1 under the premise that the heat losses from the DHW-store do not contribute to the space heating, the building has only one zone and that each kWh electricity has the same worth as one kWh pellet fuel. Under these conditions system 4 with the pellet boiler is the worst solution, mainly due to the poor boiler efficiency. Nevertheless the differences could be even higher if the higher solar gains of the combisystems did not balance it out to a certain extent and the average room temperature during the heating season would be exactly the same and not 0.3 °C to 0.8 °C lower than for the two systems with pellet stoves. System 2 needs slightly more auxiliary energy than the similar system 1. The analysis of the simulation results showed that this is mainly due to the higher total flue gas loseses in system 2.

Figure 5 shows that the total auxiliary energy and the effective store, boiler and burner losses can be reduced drastically when placing the heating system in the heated area. The effective storage losses are defined as the storage losses occurring when the room temperature exceeds 24 °С. Again, this assumes that the heat losses can be distributed effectively to the whole building. Significant are the energy savings for system 3 when operating the burner only in the heating season. In the other systems almost no effect could be seen since the stoves/ boiler have anyhow not been operated in the summer months. The effective losses in variants 2 and 3 are an indicator for the overheating problem that will occur in the summer months.

Fig. 5 Total auxiliary energy demand (left) and effective store, boiler and burner heat losses (right) of the four systems for the three variants. The dashed area represents the electrical auxiliary energy.

§5 16000 CO ~ 14000 3 12000 10000

3nt2BQtot^iiant3l

System 3 System 4

20000

18000

8000

6000

4000

2000

System 1

System 2

System 1 System 2 System 3 System 4 !□ variant 1 Dvariant 2 Bvariant 3І

The auxiliary electricity consumed has been taken into account with a conversion efficiency of 100%. More realistic for a global consideration would be to use a primary conversion efficiency of about 70% for Swedish electricity production. However, the auxiliary electricity consumption of the system is not including any parasitic electricity for pumps, controller, pellet conveyor etc. Neither are the electricity demand of the pellet heating units in included.

Table 2. Total auxiliary energy, average annual boiler and stove efficiencies, annual number of starts and stops of the pellet heater, average room temperature during the heating season and solar fraction for simulation variant 1. System 1 System 2 System 3 System 4 Unit Total auxiliary energy 17761 18061 18872 19770 kWh Average annual boiler/stove efficiency 86.9 83.3 85.0 70.8 % Number of start/stops 117 568 1201 1284 — Average room temperature during heating season 20.3 20.2 19.5 19.9 °С Solar fraction 8.6 8.1 12.1 9.5 %

The evaluation of the solar gains for such different systems is not an easy task. A very interesting method was proposed by Letz (2002) using a reference system to calculate fractional solar savings. Due to the lack of a suitable reference system and the rather

SF

Qsol

Q aux, tot

+ Qsol

(5)

This equation is assigning not all the system losses to the solar system. It expresses the ratio of solar energy supplied to the store and the total energy supplied to the store. In case of system 1 and system 2 the auxiliary electricity could be used to calculate the solar fraction for the pure solar DHW-system.

different studied system designs a simpler method was used expressed by the obtained solar fraction.

3. Conclusions

Four commercial combined solar and pellet heating systems have been investigated and evaluated according to their thermal performance particularly with regard to the heat losses of the system. The total auxiliary energy demand for heating (bought energy) is about 17800 kWh for the best system that is also the simplest system (system 1). The other systems require between 2 and 11 % more energy for heating. If the house has an open design with no obstructions for the heat transfer to ambient air system 1 is the best solution. For houses with more than one zone system 2 and 3 is a better solutions. If the whole heating system is placed within the heated area system 3 and 4 have a smaller energy demand as system 1 and 2 provided that all the losses from store and boiler/burner can unresisted be distributed in the building. For this solution the overheating problems have to be studied more closely.

All systems offer potential for improvements. An intelligent controlling of the pellet heaters would reduce the number of starts and stops. System 1 with a power modulating stove proved that although the stove was not operating on maximum power the efficiency was still better than all other systems. Further investigations are necessary to evaluate the potential of emission reductions. Moreover the interaction between boiler/burner and storage need to be improved to prevent unnecessary electricity consumption and to keep the temperature in the store as low as possible with a good stratification. The poor performance of system 4 can to a large extent be attributed to the low average boiler efficiency, which in turn is due to relatively high total flue gas losses (including leakage) and the very high boiler losses to ambient.

Flue gas and leakage losses also can be reduced by optimizing the combustion air settings or using an automatic combustion air control (lambda sensor). Leakage losses can be reduced by 6% to 25% when closing the chimney during the summer months when the burner/boiler is turned off. From 134 to 485 kWh primary energy can then be saved for system 2 to 4 per year if all leakage losses can be reduced to zero.

The solar savings for all systems are very low. With relatively simple modifications it would be possible to achieve higher solar fractions.

Acknowledgement

We are grateful to the Nordic Energy Research for their financial support for this wok within the REBUS project.