The variation in measured hot water use was also investigated on a daily basis to examine a possible difference between different days of the week. The results are found in Figure 6 for two different months the two years of available data. The trend is similar for both years and the variation between different days is small. Measured data seem to follow a similar pattern for 2005 and 2006 respectively, where the daily use is lower in 2006. Data for a certain month, however, does not coincide for the two years under investigation.

4. Discussion

The measured hot water use in the 24 apartments is similar for the two years investigated so far. The minimum in load is located in July and August, although the use is generally lower during 2006. As was also concluded in the analysis in [7] this indicates the impact of individual habits on the consumption profile. Individual habits, and the energy use in the wider context of everyday activities, can be further investigated by time diaries, which constitute the foundation of the presented model.

According to the results obtained in this study, when comparing the modelled profiles with measured, the model gives a similar, but somewhat lower hot water use. The overall load profiles, on the other hand, follow more or less the same pattern. The difference between winter time and summer time may imply that the model describes winter time better. Measurements in Malmo, Sweden, however shows morning peaks at 6-8 a. m. in weekdays and 9-10 a. m. in weekend days [9], which may indicate that the difference between summer and winter months is rather due to the routines of the particular inhabitants. The number of measured households is however too low to draw any evident conclusion.

The magnitude of the hot water use can be further compared to the study performed by the Swedish Energy Agency which shows a surprisingly low DHW use of 0.9 MWh per person and year for apartments [1]. In that study only four apartments were measured. The fact that the model predicts an energy use between that measurement study and the one investigated in this paper, and closer to the more extensive one, indicates that the model assumptions are in the right range.

The modelled profiles will be used in future studies on solar heating by introducing different profiles in dynamic simulations in the simulation tool TRNSYS. The possibility to utilize solar heat will be investigated as depending on different behaviours found in the material on time-use. Furthermore, the advantages with individual as well as joint solar heating system for a residential area will be investigated to further raise the question of tap water supply systems for low-energy buildings.

The current model only generates average profiles for one weekday and one weekend day and is not taking seasonal and weekly variations in load into account. An improvement would be to introduce those variations in the model, for example by statistical means similar to those used in Jordan et al [2]. By letting the daily individual distribution be the foundation, but vary it with day of the week and time of the year, realistic yearly profiles can be generated from single days.

More extensive measurement surveys, including a statistically significant number and distribution of households, as well as measuring hot water use in the same households as where time diaries are recorded, would be desirable to enable a more thorough validation of the model. This kind of studies are however not available at present. Although there may be deviations between the model and measurements the modelled profiles constitute an improved description of hot water use in households compared to the very simplified load profiles that are normally used.

5. Conclusion

Comparisons between modelled profiles and measurements in 24 apartments show good agreement both in magnitude and the typical distribution in time. This indicates that the model describes Swedish domestic hot water use rather well and that time-use data can be utilized for cheap and straightforward

energy estimations in households, either as an alternative, or as a complement, to detailed hot water measurements.

Detailed investigations of variations in hot water demand over the year show a clear minimum in demand during summer due to vacation periods. The average hourly distribution over the day more or less coincide between different months, and there is only a small difference in load between the days of the week, although absence and different habits seem to influence the weekly distribution between summer and autumn.

7. Acknowledgement

The work has been carried out under the auspices of The Energy Systems Programme, which is primarily financed by the Swedish Energy Agency. We also want to thank Mimer, Malarenergy, Eskilstuna Kommunfastigheter and Eskilstuna Energi och Miljo for support. Thanks to Joakim Widen, Uppsala University, for converting the original time-use data to matrixes to be used in this paper.

References

[1] Wahlstrom A, R. Nordman and U. Pettersson (2008) Matning av kall- och varmvatten i tio hushall, SP Technical Research Institute of Sweden and the Swedish Energy Agency, ER 2008:14, ISSN 1403-1892. In Swedish.

[2] Jordan U. and K. Vajen, Influence of the DHW load profile on the fractional energy savings: A case study of a solar combi-system with TRNSYS simulations, Solar Energy 69 (2000) 197-208.

[3] Wollerstrand J. (1997). District heating substations: performance, operation and design. Doctoral thesis from Department of Heat and Power Engineering, Lund Institute of Technology, Sweden.

[4] Holmberg S (1987). Flow rates and power requirements in the design of water services. Doctoral thesis from Department of Heating and Ventilation Technology, Royal Institute of Technology, Sweden.

[5] Lundh M., Wackelgard E. and K. Ellegard (2008) Design of hot water user profiles for Swedish households based on time diaries, International Conference on Green Energy with energy management and IT,

Stockholm, Sweden, March 12-13 2008.

[6] Widen J, M. Lundh, I. Vassileva, E. Dahlquist, K. Ellegard and E. Wackelgard (2008) Constructing load profiles for household electricity and hot water from time-use data — modelling approach and validation. Submitted to Energy and Buildings.

[7] Vassileva I., C. Bartusch, E. Dahlquist (2008) Differences in electricity and hot water consumption in apartments of different sizes, International Conference on Green Energy with energy management and IT, Stockholm, Sweden, March 12-13 2008.

[8] Ellegard K. and M. Cooper (2004) Complexity in daily life — a 3D visualisation showing activity patterns in their contexts. eIJTOUR (Electronic International Journal of Time Use Research), Vol 1.

[9] Bagge H. (2007) Energy Use in Multi-family Dwellings — Measurements and Methods of Analysis, Licentiate Thesis, Department of Building Physics, Lund University, Sweden.