Building load model


Loads are the most important parameters for the energy system. The reason for building energy systems is to cover the energy loads and the sizes of the loads will be the major parameters for the dimension of the system. By doubling the loads, the energy system often needs to be almost twice as large except in systems where the efficiency will increase with the size of the system.


Loads are usually separated in instant loads (power demand) and long-term loads (energy demand). In a solar energy system the power demand will usually be buffered by short-term storage, usually battery and/or a small water tank. This buffer prevents the effect from the power demand influencing the components further from the load than the short term storage. If the system is without short-term storage, the component that converts hydrogen to electricity (the fuel cell) and/or the hydrogen burner should be sized to cover the power demand.


The energy demands will influence all of the components in the analyzed system. The sizes of the solar collecting devices, the storage and the energy transforming units are influenced by the energy demand. These components require simulations over a year (or years) to find the optimum sizes.


The energy loads required by a dwelling in Norway today can be separated in several categories, such as electricity, high temperature heat, low temperature heat and sometimes cooling. Electricity can be considered the most valuable form of energy in this system. This is because it can be used to cover all demands. This will be true in all systems as electricity can be converted to both high and low temperature heat (or cooling) without or with small losses compared to the other way around.


The electrical load will supply the energy needed by lighting and electrical appliances (TV, microwave, stereo, washing machine, fans etc.). The high temperature demand is energy needed for cooking, and it is supplied by hydrogen burners or electricity depending on what system is chosen. The low temperature heat includes the need for hot tap water and space heating, and it can be supplied by waste heat from system components, solar collectors, hydrogen burners or also by electricity. For Norwegian conditions, cooling is rarely needed and is almost never available in dwellings. Consequently, there is no cooling load included in the simulations.


The load sizes are mainly taken from the report by Jacobsen, Raaen and Hestnes, 1988 for a low energy dwelling in Trondheim. The loads are 5540 kWh for electricity, 730 for high temperature heat (written as an electrical load in Table 3.1.) and 6450 kWh for low temperature heat per year. Table 3.1. shows the load in detail.






















Misc. equipment



Fans, pumps



Domestic hot water




Heating of dwelling









Peter Lund, 1994 has also made some calculations for the technology of today and tomorrow. The loads for 60o calculated by Lund are given in . These calculations of the loads can be used to evaluate a future system low energy houses.


The distribution of the electrical and high temperature loads over the day is assumed to be similar to the description made by Mørner, 1989 with the major part of the load in the morning and afternoon. The loads are not correct for periods the inhabitants usually will be gone (for example summer and Easter vacations) and a different use pattern during the weekend.

60o North

Heat [kWh]

Electricity [kWh]

Standard efficient house



Best available house



Future low energy house




Besides the loads required by the user, the system itself will use some of the energy output. This load is called parasitic loads and can be both thermal and electrical. Loads that are considered parasitic are for example: electricity required by controllers, electricity to run pumps, fans etc. and heat to (pre-) heat the components or parts of the component system.



1. Thermal load



1.1. Space heating


Neither of the space heating loads is modeled in great detail. The reason for this is that the emphasis in this work is on the PV- hydrogen system. A very detailed building model can be very time consuming both to make the description of the dwelling and to do the simulations. Therefore, the building is only modeled approximately and the energy consumption is made for one "typical" day that is used throughout the year. The solar gain is only approximated as well as internal gains and gain from equipment. Ventilation is not simulated, but the heat loss from ventilation is included in the heat loss factor and the energy demanded by fans and pumps are included in the electrical load.


The indoor temperature is assumed to be at least 20 oC but it is allowed to float up (to avoid any cooling load to be included in the simplified TRNSYS simulation of the building load). The cooling of the house is usually no problem in Norway, since the ambient temperature anywhere in Norway rarely exceeds 30 oC and only some days in summer the temperature exceeds 20 oC. This means that cooling of dwellings can be done by opening windows through most of the year.


The heat loss is calculated hour by hour from the heat loss factor, ambient temperature and the indoor temperature. To do the simulations of the heat load, the gains and the heat loss factor have to be estimated.


Solar energy gain

The solar energy gain comes through the windows directed east, south and west in the house. This will reduce the heating load only when there is a heating demand due to a low heat capacity of the house. The highly insulated house will mainly have a heating load during winter, late fall and early spring. During this time the solar radiation will mainly come from the south and through the south facing windows. The south facing windows will for that reason have the biggest influence on the space heating load.


The total space heating load for the low energy dwelling is calculated by Jacobsen et al., 1988. An increase in window area will cause higher solar energy gains. The total heat loss factor from the dwelling will then be increased so that the yearly space heating load stays constant and equal to what was reported by Jacobsen. This means that a change in window area will shift the load but not decrease/increase the total space heating load. A greater window area will shift the load towards winter and decrease the space heating load in spring and fall. The shift in heating load per area window will be reduced with increasing window area. This is because less of the increasing solar gain will be useful with larger window area.


The solar gain will also influence what time of day the heating load will come. In the TRNSYS simulations, the distribution over a day will influence the efficiency of some components and how the energy can be utilized.


The windows are assumed to directly face the south, east and west. If the house does not face these directions, the area facing the east/west/south can be calculated by simple trigonometric calculations (as an estimated method) or an adjustment in the angle of the windows can be done in the TRNSYS deck.


Another factor that will influence the amount of solar gain over a day is that the indoor temperature is allowed to float. This will mean an increased "storage" of solar energy compared to the real dwelling where the inhabitants will start to open windows when the indoor temperature is high. The increased temperature range that the mass of the dwelling is varying between, will shift the space heating load towards winter. The heat capacity of the low energy dwelling was not available and estimated to be 1000 Wh/K (10 kJ/m2K (light walls, floors, ceilings, (Kolsaker et al., 1993), 142 m2 floor area => 3*142*10 ~ 1 kWh/K). This heat capacity might be a little low because of for example concrete floors, but the whole interior building mass will probably not be used as heat storage. Only the part of the mass in rooms with windows facing the south, west and east is used as heat storage. The effect of the increased storage by the floating temperature will be reduced or eliminated with a lower heat capacity of the house.


Energy gain from humans

It is assumed that a family of four inhabit the dwelling. At night the inhabitants will approximately give off about 40 W for the children and 75 W for the grown ups. At day time they will be more active and the heat from the inhabitants will be 70 W and 125 W respectively.


The family will start to get up at six a.m. and the whole family have left by 9 a.m. They start to get back at 1 p.m. and all the family members will be home from 4 p.m. to 6 p.m. From 6 p.m. to 8 p.m. it is assumed that half the family is home and the rest of the evening the whole family is home, and they go to bed at 11 p.m.


Energy gain from equipment and light

All the energy used as electricity is transformed into heat and most of this heat will be internal heat gain. The exception is some light that is assumed to be outside, and 70% of the heat from cooking that is assumed to be vented out. The heat gain from light is assumed to be constant over the year and derived from the winter load to lighting (see DC-load in Figure 3.5). The internal energy gains are shown in Figure 2.


Heat loss factor

To achieve the total load over the year for the dwelling, the heat loss factor is set to 85 W/K (if 3 m2 window area in each direction is assumed), due to the possibility to add the gains from solar radiation through windows. The load simulations are using the weather generated by TRNSYS from monthly data. By modeling the house this way, the dwelling can be simulated in several climates without changing the input deck.



1.2. Tap water


It is assumed that the tap water need is constant over the year and week. The tap water consumption is assumed to be 10 kWh per day, that is 3650 kWh per year. The main load is assumed to occur in the morning and to a smaller degree in the afternoon/evening. However, this load is leveled by the hot water tank.



3.1.3. Cooking


The cooking load is a thermal high temperature. This load can be covered either by electricity or by burning hydrogen in a catalytic or other kind of hydrogen burner. The most efficient way of covering this load will be by burning hydrogen in winter and by using electricity in summer. This is because electricity is available directly from the PV- panel in summer while hydrogen is the only energy source during winter.


The most efficient solution will however require two stoves, one electrical and one to burn hydrogen. This might not be very practical and a choice of stove will usually have to be made. The hydrogen burner will be more efficient during winter than an electrical stove due to a higher loss in the fuel cell compared to the hydrogen burner. If a hydrogen burner is used during summer the electricity has to be converted into hydrogen in the electrolyzer and then burned. An electrical stove will usually be more efficient during summer due to the utilization of electricity from the PV panel and the battery. The difference over the year might not be very large if most of the "waste" heat from the fuel cell can be used for other purposes (space heating and tap water).



2. Electrical loads


The electrical load needed for simulations in TRNSYS is split into a DC-load and an AC-load. This is done because a DC/AC converter for the AC load will decrease the energy efficiency of the system. The DC-load will only include the lighting demand, while the rest of the electrical load is assumed to be AC-load (in the future more of the components can probably run directly on DC



2.1. Direct current (DC)


The direct current load is assumed to be the lighting. During night it is assumed outdoor lighting and/or lighting of a hall. Some rooms will not have day lighting and light will be used in these rooms even during day. Due to this, the lighting load will have a minimum level of 50 W.


At times when the predicted lighting load is higher than 50 W and it is light outside, the lighting load will be reduced to 50 W. This will cause a slightly sinus formed curve over the year. For the low energy dwelling in Trondheim the lighting load is 1400 kWh/year.



2.2. Alternating current (AC)


The alternating current will cover the load from all other electrical components than light. The AC load is assumed to be constant every day over the year. For the low energy dwelling in Trondheim, this load is 11.3 kWh per day or 4140 kWh/year.