Measuring drinking water temperature changes in a distribution network

Abstract

During distribution the drinking water temperature changes due to the temperature of the soil surrounding the drinking water distribution network (DWDN). A drinking water temperature below 25 °C at the tap is required to meet Legionella prevention standards and/or drinking water standards. With climate change, urbanisation and the energy transition towards more district heating networks, the urban subsurface will heat up further, and more exceedances of the 25 °C threshold are expected.To understand the effectiveness of various measures to keep drinking water temperature below the threshold a modelling approach was followed. The drinking water temperature model (WTM) calculates drinking water temperatures at each customer from a hydraulic network model and heat conduction from the outside of the pipe wall, where the soil temperature is kept constant and the soil is thus assumed to be an infinite heat source. The WTM was validated with measurements in a DWDN on relatively small diameters (150 mm or smaller) and at locations far enough from the source so that equilibrium with the soil temperature was already reached (i.e. the influence of residence time was not validated). In reality, the soil is not an infinite heat source, but is affected by the drinking water temperature. We therefor developed an enhanced WTM (called WTM+) which uses an extra insulation layer (of soil material) around the drinking water pipe to account for the soil which is affected by the drinking water temperature. In order to determine a suitable length scale for this extra insulation layer, and to validate the WTM+, we measured drinking water temperatures in two real case studies.Case study 1 is a single 1 km pipe where we manipulated flows, and thus residence time. Case study 2 is a DWDN with a variation in diameters and residence times. Designing and executing these measurements proved to be quite a challenge. In case study 1 the incoming drinking water temperatures and soil temperatures were not stable during the three weeks of measurements. The temperature changed typically less than 1.5 °C over 1 km, and the sensors had a resolution of only 0.1 °C. In case study 2 the measurements were done on a single day (morning, and repeated in the afternoon), and were quite stable. The drinking water temperature changed up to 8 °C over the course of the residence time. However, as the DWDN hydraulic network model is never perfect, the residence times are not all known accurately, and the surrounding soil temperatures, that may have varied quite a lot over the DWDN, were not measured. Nevertheless, the case studies did prove to be suitable for validating the WTM+, including the effect of residence time.