Commercial refrigerant industry is one of the biggest contributing factors toglobal warming according to a study by Mota-Babiloni et al. (2015). Studies have shown that emissions of fluorinated greenhouse gases' effect on global warming can increase from 1.3% (2004) to 7.9% (2050) if no change in behaviour is implemented in the industry (Francis et al., 2017). At present, the F-Gas Ordinance sets requirements for leakage warning systems and periodic leakage detection.The most common leak detection systems consist of sensors locatedoutside the cooling process and measure the amount of coolant in the air. However, there are limitations with sensors, including measuring leaks in engine rooms with high air circulation and that the sensors are unable to measure small leaks and mistake other contaminants for refrigerant leakage (Leonardsson, 2021).The purpose of this study is therefore to investigate where leakage occurs in a refrigerant process, howoperating data changes andto propose an alternative detection method for refrigerant leakage.Johnson Control has other technical solutions such as their Performance Index which compares a cooling process operation in practice compared to the theoretical performance that the unit should deliver. In their own test facility, however, it was discovered that a unit can lose up to 20% of its refrigerantmassbefore a clear reduction in COP was discovered(Thorén, 2021).To discover where the leakage in a cooling process occurs, a field study was conducted on a cruise ship docked in Barcelona. The first field study on leak detection was initially carried out by direct method, with the help of an electronic leak detector. The indirect methodwas also used to complement the direct method, this was doneby ocularly looking for oil stains on and around the cooling process. It showed that leakage almost always occurs on the high-pressure side of gaskets, couplingsand plugs. Similar discoveries were made inthe study byFrancis et al. (2017).The results also showed thatthemore connectionsthere are on a unit, such as plugs or service valves, the greater probability that leakage will occur. A solution to reduce leakage on units with many connections,such as supply systems, couldbe to adjustthe design of the systemsand simply remove plugs for connections that are no longer usedfor different reasons.A second field study was conducted on the same cruise ship to determine what operational data changesdue to refrigerant leakage. The unit was emptied into a service tank for two minutes while operating data was logged. After analysis of the results in 16 different operating data points, three of themwere selected for the alternative leak detection model, the DGS-model.Data for longer operating periods, approximately 5 months, with fault-free operation for the selected parameters Θoutin evaporator, expansion valve opening degree and suction   overheating were compared with the operating data from the leakage experiment in field study 2. Thesethree were selected from a total of 14measured operating datapoints. The simultaneous increase ofall three operating data points indicatesrefrigerant leakage. That fact is the major reason for the selection of the three operating data points.Aprobability calculationbased on the three data pointswas created to calculate the probability of thedata points being outside a given range. If the probability calculation says that the chance of the reported value is zero, it is indicated that there is a problem with the process, most likely a refrigerant leak.As a proposal for further work,our proposal is to carry out validation tests of the DGS model on units in operation and a more accurate experiment on refrigerant leakage with operational changes per removed mass.

Modern Heavy Duty Vehicles consist of a multitude of components and operate in various conditions. As there is value in goods transported, there is an incentive to avoid unplanned breakdowns. For this, condition based maintenance can be applied.\newline This thesis presents a study comparing the applicability of the data-driven Consensus SelfOrganizing Models (COSMO) method and the model-driven patent series introduced by Fogelstrom, applied on the air processing system for leakage detection on Scania Heavy Duty Vehicles. The comparison of the two methods is done using the Area Under Curve value given by the Receiver Operating Characteristics curves for features in order to reach a verdict.\newline For this purpose, three criteria were investigated. First, the effects of the hyper-parameters were explored to conclude a necessary vehicle fleet size and time period required for COSMO to function. The second experiment regarded whether environmental factors impact the predictability of the method, and finally the effect on the predictability for the case of nonidentical vehicles was determined.\newline The results indicate that the number of representations ought to be at least 60, rather with a larger set of vehicles in the fleet than with a larger window size, and that the vehicles should be close to identical on a component level and be in use in comparable ambient conditions.\newline In cases where the vehicle fleet is heterogeneous, a physical model of each system is preferable as this produces more stable results compared to the COSMO method. / Moderna tunga fordon består av ett stort antal komponenter och används i många olika miljöer. Då värdet för tunga fordon ofta består i hur mycket gods som transporteras uppstår ett incitament till att förebygga oplanerade stopp. Detta görs med fördel med hjälp av tillståndsbaserat underhåll. Denna avhandling undersöker användbarheten av den data-drivna metoden Consensus SelfOrganizing Models (COSMO) kontra en modellbaserad patentserie för att upptäcka läckage på luftsystem i tunga fordon. Metoderna ställs mot varandra med hjälp av Area Under Curve-värdet som kommer från Receiver Operating Characteristics-kurvor från beskrivande signaler. Detta gjordes genom att utvärdera tre kriterier. Dels hur hyperparametrar influerar COSMOmetoden för att avgöra en rimlig storlek på fordonsflottan, dels huruvida omgivningsförhållanden påverkar resultatet och slutligen till vilken grad metoden påverkas av att fordonsflottan inte är identisk. Slutsatsen är att COSMO-metoden med fördel kan användas sålänge antalet representationer överstiger 60 och att fordonen inom flottan är likvärdiga och har använts inom liknande omgivningsförhållanden. Om fordonsflottan är heterogen så föredras en fysisk modell av systemet då detta ger ett mer stabilt resultat jämfört med COSMO-metoden.