12th International Symposium on District Heating and Cooling

4. SELECTION OF INSULATION THICKNESSThe reason for the relatively high heat losses for thetwo smallest diameters is explained by Fig. 5.Thickness insulation [mm]PB std insulation thickness454035302520151050PB extended insulation thickness16 20 25 32 40 50 63 75 90 110Nominal diameter [mm]Fig. 5, Insulation thickness, standard and increasedIn red, this graph contains the insulation thickness ofthe current PB-PE-PE product range. The somewhaterratic distribution of insulation thickness over therange is caused by the use of customary dimensionsfor the outer casing. It can be seen that for the twosmallest diameters the insulation is rather thin, whichexplains the relatively high heat losses.In purple, the graph in Fig. 5 shows a modified range,with increased insulation thickness for some of thesmaller diameters, as in general it is easier to achieve agood insulation quality for the smaller dimensions. Theheat losses of the modified range were calculated andare presented in Fig. 6.Heat Loss [W/m]50454035302520151050St stdPB std insulation thicknessPB extended insulation thickness16 20 25 32 40 50 63 75 90 110Nominal diameter [mm]Fig. 6, Heat loss per pair, including increased insulationthicknessThe <strong>12th</strong> <strong>International</strong> <strong>Symposium</strong> on District Heating and Cooling,September 5 th to September 7 th , 2010, Tallinn, Estonia312The graph in Fig. 6 shows that the modified range (inblue) is much closer to the reference (Steel-PUR-PE, inpurple). Heat loss is (almost) proportional to thediameter, which seems about right. However, themodified range of insulation thickness in Fig. 5 (purple)still shows a somewhat erratic distribution, whichsuggests that the current range of customarydimensions for the outer casing does not lead to anoptimal distribution of insulation thickness. It may proveworthwhile to develop a new range of outerdimensions, adapted to the diameter of the mediumpipe.5. IMPROVEMENT OF INSULATION QUALITYDue to the new testing facilities described by vanWijnkoop et al. in [3], the process of productimprovement has been speeded up considerably.During the course of the investigations, resulting in thispaper, it is becoming clear that further improvement ofthe insulation quality is feasible. The measurementprinciple used for the determination of heat losses doesnot allow for direct measurement of the insulationproperties of the foam; however, some sort of―equivalent lambda‖ can be derived from the data bycalculation. As explained by van der Ven et al. in [4],insulation quality differs for different diameters. Forproduction reasons, it is not expected that insulationquality will reach the same level over the entire productrange. Typically, the higher values will be reached inthe smaller dimensions. Still, an educated guess canbe made as to which levels are feasible from atechnical viewpoint, see Table 1.Table 1, Improved insulation quality, ―equivalent‖ or―synthetic‖ lambda at 50 °C mean temperatureType Area Lambda fresh Lambda Degassed50A25 1074 0.0283 0.032663A32 1701 0.0287 0.033075A40 2364 0.0291 0.033590A50 2993 0.0295 0.034090A40 3670 0.0300 0.034590A32 3886 0.0301 0.0346125A63 6204 0.0316 0.0363160A90 10790 0.0345 0.0397160A75 12611 0.0357 0.0411200A110 16879 0.0385 0.0442Please note: The lambda values in Table 1 are not themeasured lambdas of samples of the insulation foam,and may not be interpreted as actual physicalproperties of the insulation material. The values werecalculated on the basis of heat loss measurements ofsections of pipe according to EN15632, and thereforeare some sort of ―synthetic system lambdas‖.