**5. Discussion**

The simulation results for the "natural gas scenario" show some interesting conclusions. In this scenario, natural gas is used as the only fuel for all CHP systems. Fuel cell-based technologies reform the gas to obtain pure hydrogen, which results in an additional gas consumption. Even so, the fuel cell-based CHP systems have a significant impact into PE reduction as can be observed in Figure 7a1,a2. It is noteworthy that the smallest fuel cell CHP ("1 kWe FC") system presents the greatest PE reductions for the 10-dwelling building simulations. PE consumptions are from ca. 50% for B-Id 1 to ca. 90% for B-Id 4. On the contrary, the same facility does not show such good behavior in the single dwelling simulation. This is due to the better efficiency of the back-up heater for the energy conversion of natural gas into heat. In other words, the "1 kWe FC" system used in the high-rise buildings means a longer operation time because of the lower proportion of heat demand fed from the CHP (see Figures 11b and 12b for operating time and thermal demand respectively). Thus, more energy from the back-up heater is needed to cover the heating demand and this means that lower energy from the natural gas is required. Even so, the use of the CHP decreases the consumed energy from the electrical grid, which benefits the PE reduction. This can also be observed in the CO2 emissions in Figure 8a1,a2, where only fuel cell-based CHP systems show an effective greenhouse gases reduction because of the impact of the reduction in electrical energy importation. It is easy to see in Figure 7a2 that the higher the HtP ratio the better the PE reduction.

"MICAPEMrated" fuel cell CHP was designed according to a standard dwelling consumption, showing the best PE reduction in all scenarios. Focusing on the first scenario (Figure 7a1) the PE consumption is around 80% for the user with the higher HtP ratio (Id 1) and around 90% for the lower HtP ratio (Id 4). However, CO2 emissions are maintained because of the increment in natural gas consumption to meet the thermal demand (Figure 8a1). Electrical energy surplus achieves a maximum value, doubling the energy demand, as can be observed in Figure 9a. In the same Figure, Id 4 and B-Id 4 always import electricity from grid because the datasets are based on a high electrified user with a low HtP ratio, ca. 0.7.

The results for the "green gas scenario" are also interesting and as expected, the use of carbon-free fuel has a significant impact on the PE reduction in buildings. The higher the energy exported; the higher PE reduction is achieved. Main differences for fuel-cell-based systems in Figure 7b1,b2 are the PE reduction rate sign. On the one hand, in Figure 7b1 it is negative because the CHP unit is able to cover the entire thermal demand (see also Figure 11a) yielding an exported electrical energy surplus. On the other hand, 10-dwelling building results in Figure 7b2 show that the reduction is not so high because neither the thermal nor the electrical demand are fully covered by the CHP unit and despite the fact that green gas can be used in the backup heater, the electricity has to be imported from the commercial grid. Decreasing the energy importation rather than creating an energy surplus is the preferred situation for the actual electrical systems because energy consumers can be managed more easily than small energy producers (self-consumption without net balance). The case Id 4 is an exception not only in the carbon-free fuel scenario, but also in the other ones, because it is a highly electrified consumer with a low thermal energy demand. In fact, this means that the total operation time of the units is lower than others, so electricity has to be imported. Nevertheless, the PE reduction is significant for Id 4 due to the cogeneration.

The "expected scenario" is the most probable scenario, where only hydrogen is used to power the fuel cell-based CHP systems and natural gas for back-up heaters and the other CHP technologies. This means that the fuel consumption term in Equation (6) can be neglected for fuel cell-based system exclusively. Figure 7c1 shows that the most powerful thermal CHP units ("gas ICE" and "gas turbine") result in a worst PE reduction for all dwellings due to the low operating time (see Figure 12a) and, consequently, the lowest electrical energy production (see Figure 9a). The negative value in the PE ratio for fuelcell-based CHP units is due to the electrical energy surplus. Figure 7a1 shows, again, an optimal design point in the "MICAPEMrated" characteristics. For the 10-dwelling buildings, the higher the power the higher the PE reduction due to the ability to provide the energy demands from a low consumption of carbon-free fuel for PE production. The contrary can be observed in Figure 7a2 for the "natural gas scenario", where a greater fuel consumption is penalized.

Figure 8a1,a2 show the CO2 reduction results for the "gas-fueled scenario" for both individual and building dwellings, respectively. The use of CHP in dwellings does not have a carbon emissions reduction due to the higher gas consumption because of the efficiency reduction compared to the use of a boiler. Nevertheless, the 10-dwelling building simulation shows that CO2 emissions are lower for the fuel cell-based system due to the electrical generation and the increase in global efficiency. Similarly, in both the "green fuel scenario" and the "expected scenario", the reduction of CO2 emission shows the same behavior compared to the PE reduction.

The size of the CHP system is an important design and a critical economical parameter. The size of the CHP is directly related to the ability to meet the energy demand, but indirectly related to the operation time (and the economic viability). Figure 11 shows the share of thermal energy demand produced with the CHP system and Figure 12 the share of operation hours per year. The total operating time for the smaller units of the "*gas ICE*" and the "gas turbine" CHP units applied to low thermal demand consumers like a single dwelling (Figure 12a), makes its use unviable (less than 5%-year hours). "Gas turbine" units still are unviable for typical buildings in Spain (less than 20%-year hours). Contrary, fuel cell-based technologies, due to their lower power appear to be a better solution for CHP systems in the building sector, ca. 80%-year hours in the best cases. The low thermal power is not a handicap because fuel-cell-based CHP systems are fully scalable.
