Low temperature radiator heating distribution and emission efficiency in residential buildings

Abstract Low and nearly zero energy buildings with decreased heating need can utilize low temperature heating systems for energy efficient heating. Distribution and emission losses for low temperature radiator heating cannot be found from European standards or scientific literature. The use of the losses of conventional systems can result in significant overestimation of heating energy use. In this paper, distribution and emission losses of low temperature and conventional radiator heating system were determined in North and Central Europe climates for low energy detached houses and apartment buildings. Detailed dynamic components of heating system in the whole building energy simulation model allowed to quantify these losses. Main findings of the study show that distribution losses can be controlled with low temperature heating curves and emission losses with PI type thermostats. For conventional systems the losses higher than 50% of heating need were calculated in the apartment building. With low temperature heating curve, PI thermostats and limited heating period distribution and emission losses were possible to keep below 1% in detached houses in both climates. In apartment buildings the minimal achievable losses were significantly higher, between 6 and 12% in North and Central European climates, respectively. Proportional thermostats add 2 to 6% to these losses. Based on results, heating curve of 45/35 °C can be recommended for detached houses and even 40/30 °C for apartment buildings. Insulating distribution and connection pipes in heated spaces proved to have no practical effect on heat losses. Compared to EN 15316-2-1:2007, the losses were significantly lower especially for low temperature heating curves. A new set of tabulated values is proposed for the revision of the standard. It was possible to explain the mechanism of losses, but still it was not possible to calculate losses from hourly energy need data with the correlation equation leading to overestimation by factor of about 10 compared to dynamic simulation.