12th International Symposium on District Heating and Cooling

The <strong>12th</strong> <strong>International</strong> <strong>Symposium</strong> on District Heating and Cooling,September 5 th to September 7 th , 2010, Tallinn, Estoniademonstrated [6] [3]. Euroheat and Power recommendthat a 2-stage connection be used only in large multiresidentialbuildings if the primary radiator returntemperature is high. However, it should not beemployed if a low-flow heating system providing lowreturn temperatures is used [1].The advantage of extending the HEX when thesecondary flow is low actually demonstrates theoptimisation problem: When the secondary flow isreduced, the secondary return temperature willdecrease. In the radiator HEX, the situation is different.As the secondary flow decreases, the differencebetween the primary and the secondary returntemperatures, increases as a result of the heat transfercoefficient in the HEX being strongly flow dependent.Fig. 2 shows how the secondary return temperature islowered with a decreasing secondary flow while thedifference between primary and secondary returntemperatures increases. This results in a primary returntemperature that, at first, decreases and then increaseswhen the secondary flow is further reduced. The valuesin the figure have been taken from one of the testobjects. For this heat load, the lowest primary returntemperature was achieved for a secondary flow ofapproximately 30 % of the original flow.Return temperature, °C4039383736353433Tp,r,radTs,rGrädigkeitOptimum, lowest T p,r15 20 25 30 35 40 45 50Flow, %Fig. 2 Primary and secondary return temperatures, as wellas the difference between them, as functions of theradiator flow.Another reason for including the impact of an extendedHEX in the comparison in Table 1 is the opportunity ofconnecting to new installations. Large parts of thehousing stock in Sweden, built under strong politicalincentives during the 1960s and 1970s, are facingsubstantial renovation needs. The results of this projectcan be considered consistent even if fewer radiatorsystems be oversized in the future, whetherincorporated in older, renovated, or new buildings. Thesmaller potential for return temperature reductionsresulting from less oversized radiator systems may becompensated by the ability to install a HEX that isdimensioned for of an optimised radiator programme,76543210Grädigkeit, °Ci.e., a longer HEX. Furthermore, with optimised control,there exists a preparedness for future changes insystem temperatures in the DH network. Should theDH supply temperature be changed, an adaptivecontrol will ensure that the lowest possible returntemperature is always achieved.In order to operate according to Fig. 1, the algorithmmust combine a control of the radiator supplytemperature with a control of the radiator flow as afunction of the heat load and the DH supplytemperature. In previous work [7], we have shown thatit is possible to manually determine the optimal radiatorsupply temperature and flow. A natural continuation isto develop a method for automatic adjustment ofparameter values for the optimal control algorithm.THE TEST OBJECTSThe tests have been carried out in four multi-residentialbuildings in the city of Karlshamn, Sweden. The houseswere built in 1967-1968: three of them had three storiesand a basement, and one had six stories and abasement. The number of flats varied between 20 and30 per house.The radiators in all houses were fitted with TRVs, butthese were at least ten years old. It was thus uncertainwhether they functioned properly. The circulation flowwas found not to vary significantly in any of the radiatorcircuits, which may have been an indication that manyof the TRVs were not working. However, it should benoted that the presented control algorithm isindependent of the use of TRVs in a system. Whatevercombination of optimal supply temperature and flowthat is identified for a given outdoor temperature, theheat supply will be the same. The main task for TRVsis to limit the heat supply in a room where additionalheat supply (solar radiation, bodily warmth or electricalequipment) would result in an overheating of the room.The substations were of the 2-stage type and equippedwith control logics of the brand IQ Heat (Alfa Laval AB).The equipment for the building automation wasmanufactured by Siemens and furnished with aseparate communications module that could also beused for executing minor computer programmes. Therewas also an internet connection, rendering it possibleto communicate in a number of ways, such as via thesoftware Saphir ScopeMeter© (Siemens), or FTP. Aftera reconfiguration, the pump speed could be controlled,since all pumps were equipped with communicationmodules.In order to monitor the circulation flow in the radiatorcircuits during the tests, clamp-on ultrasonic flowmeterswere utilised. However, the objective was todevelop a control algorithm based on modern, state-ofthe-artequipment without using additional installations.208