Beta 1


Title Design af Standardløsninger til Energibesparelser i Industrien
Author Ilsøe, Jakob
Kristensen, Kenneth Tolstrup
Supervisor Qvale, Einar Bjørn (Institut for Mekanisk Teknologi, Danmarks Tekniske Universitet, DTU, DK-2800 Kgs. Lyngby, Denmark)
Elmegaard, Brian (Termiske Energisystemer, Institut for Mekanisk Teknologi, Danmarks Tekniske Universitet, DTU, DK-2800 Kgs. Lyngby, Denmark)
Dalsgård, Henrik (Cowi)
Institution Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark
Thesis level Master's thesis
Year 2004
Abstract Implementing process integration in the industry can be time consuming and complicated. Therefore, it has been recommended to concentrate on simple and standard "off the shelf"-solutions. The advantages of standard systems compared to custom made systems are that these can be designed quickly and that the heat recovery network, which is generated, is reliable. The duplication of simple network structures also makes it possible to reuse the operation and maintenance procedures and to benefit from the already acquired experiences. An additional advantage is that recommendations from satisfied users make it easier to convince managers of other plants to adopt the technology and implement the heat recovery solutions. Based on these experiences and conclusions in H. Dalsgaard’s Ph.D. thesis from 2002 a simulation tool was developed using the WinDali modelling package. The standard solutions contained in the model are entirely based on heat recovery from cooling plants and in this way the basic structure in the model is divided into two parts. The plant can operate with a maximum of three different evaporating temperatures. A key feature of the model is the opportunity to specify cooling demand profiles for each temperature level, each profile covering a period of 24 hours. Thus, the model takes into account the impacts on heat recovery from variations in production schemes. Based on an investigation of part load characteristics of compressors, it was decided to implement these in the model. As these depend up the method of control, the user must specify whether part load is achieved by using frequency converter or not. At all cooling demands the operation of the compressors is calculated according to a maximum Coefficient of Performance criteria. The required cooling is supplied by up to four compressors at each temperature level, and all compressors in the plant compress to the same condensing pressure. A total of 16 reciprocating compressors and 14 screw compressors manufactured by York Refrigeration have been added to the list from which the compressors are to be chosen. The screw compressors can be operated with or without an open economizer. Furthermore, the inlet temperature of the oil to the screw compressors can be set. Four standard heat exchanger networks for heat recovery were implemented. Heat recovery in each heat exchanger is calculated by applying the Log Mean Temperature Difference and the first law of thermodynamics. In all networks the heat is used for warm water production. All four networks include direct recovery of waste heat from the desuperheater and a condenser. Under normal circumstances, it is only possible to recover a fraction of the total condensing heat and therefore an evaporative condenser was implemented to the networks. Networks one and two vary in respect to the utilization of oil cooling which is only included in network one. Networks three and four include a heat pump in order to increase the amount of produced energy, but oil cooling is only utilized in network three. The modelling of the heat pump was m based on an actual heat pump, also manufactured by York Refrigeration. In the heat recovery systems, key parameters are heat transfer coefficients and temperature levels. After running a simulation, temperatures and heat transfers in the heat recovery system can be visualized as a function of time by graphs. Additionally, integrated key results, such as compressor power consumption, heat recovered, and overall production of water are displayed. Using this model, the cooling plant at the Danpo poultry processing plant in Farre, Jutland, was used for a case-study. After analysing the cooling plant and the heat recovery system at Danpo (first network) simulations were conducted in order to compare calculated results with actual data. The comparison proved a satisfactory compliance with a discrepancy of 7-18% in respect to the produced amount of warm water. Some deviations were found when comparing simulated temperature levels with observed temperatures at the plant. Changes in the screw compressor oil temperature were likewise found to influence the amount of produced warm water. Decreasing the temperature by one degree Celsius results in additional 0,18m3 warm water at Danpo. Care should be taken when decreasing the oil temperature, as it causes a drop in the upper limit of how warm the water can be produced. Aiming at more heat recovery from the Danpo plant, simulations with the use of a heat pump were conducted. By this method the simulated warm water production was increased from a normal production of 92m3 a day to as much as 190m3. However, the simulations showed that a increase in the heat transfer coefficient of the desuperheater was necessary to uphold the temperature levels. For a production of 170m3 a day the costs of implementing the heat pump was estimated to 500,000DKK and the yearly savings to 170,000DKK. Provided the need for water is present, the use of a heat pump in the heat recovery systems seems attractive.
Imprint Technical University of Denmark (DTU) : Kgs. Lyngby, Denmark
Admin Creation date: 2009-11-05    Update date: 2010-10-28    Source: dtu    ID: 252068    Original MXD