Målet med den här rapporten är att undersöka hur olika systemjordningar påverkar felström och nollpunktspänning vid enfasiga jordfel. Den praktiska delen består av att konstruera en laborationsuppställning där experimenten ska göras. För personsäkerheten kommer den att köras på 50 V huvudspänning. Den kommer innehålla 5 olika delar. Den första delen är inkommande 400/50 V-transformator, andra delen är själva jordningsmodulen, tredje delen är fördelningen, fjärde delen är kabelmodellerna och till sist är det en felmodell där olika typer av fel kan simuleras. En kort laboration kommer att utföras på uppställningen där enfasiga jordfel med olika felresistans kommer att skapas och mätas på. Den teoretiska delen består av att göra beräkningar på de olika systemen för att se hur bra laborationsuppställningen stämmer överens med det teoretiska. Den kommer även innehålla en del om hur allt detta kan överföras till ett 10 kV-system. Resultatet visar att det praktiska och teoretiska stämmer ganska väl överens, men särskilt i fallet med det spoljordade systemet är det svårt att bestämma de verkliga fysiska egenskaperna för att kunna göra tillräckligt noggranna antaganden för beräkningarna. I jämförelsen mellan de olika systemjordningarna blir deras olika egenskaper uppenbara. I ett nät med kompenserad nollpunkt blir t.ex. jordfelströmmen vid lågohmiga jordfel betydligt lägre än för de andra typerna av nät. Det ger även en högre nollpunktspänning vid högohmiga jordfel. / The aim for this report is to compare different earthing systems in electrical networks and examine how the fault currents and neutral point voltages depends on the earthing system during single phase to earth faults. The practical part consists of constructing the laboratory model where the experiment will take place. For personal safety the model will be using a main voltage of 50 V. The model will consist of 5 modules. The first module is the mains 400/50 V transformer, the second is the earthing module (where different types of earthing can be chosen), the third module is the distribution block, the fourth module is 3 cable models and the fifth module is the fault model where different types of fault can be simulated. After the construction a short experiment will be carried out where earth fault with different resistances on networks with different earthing system will be created and measured upon. The theoretical part will consist of calculations verifying the created model accuracy with the theoretical model found in literature. There will also be a short part explaining how all this applies to a 10 kV system. The results from this is that the created model and the theoretical model do comply but in the case of the system with a Petersen-coil it is hard to determine all the real physical properties of the system to make accurate assumptions for the calculations. Comparing the different types of earthing it is clear that in a system with a Petersen-coil the earth fault current with a low impedance fault is much smaller than for the other types of systems. As for the neutral point voltage it is higher in a system with a Petersen-coil when there is a high impedance fault compared to the other types of earthing systems.

The rural power grid has traditionally mostly consisted of overhead power lines. In recent years the trend has been to replace the overhead lines with cables instead. The reason is that overhead lines are relatively vulnerable, strong winds and storms can cause trees and branches to fall over the power lines and cause a phase to ground fault. This will then trip the ground fault relays and disconnect the faulty power line. A cable grid is not vulnerable in the same way, and could be considered a solution to make the power grid more reliable. A cable grid does come whit other types of problems instead. It generates about 50 times more phase to ground capacitance compared with the same length of overhead lines. When a phase to ground fault occurs the capacitance in the healthy phases will generate a current to ground and then through the fault. On average a cable grid generates about 2 A per kilometer. Large cable grids can therefore cause very large capacitive currents to flow through the fault.  To counter this, a reactor is placed between the neutral point of the transformer and ground. When a phase to ground fault occurs, the reactor will generate an inductive current which is in the opposite phase compared to the capacitive current. This current will flow through the faulty line and cancel out the capacitive current. In a perfectly tuned power grid the only component left in the fault is a smaller resistive current. Large cable grids will require a large reactor to generate the large inductive current, which might need to flow over a great distance in the grid to reach the fault location. To reduce the inductive current from the central reactor, it is possible to install smaller local reactors in the grid. These will then in the event of a phase to ground fault generate a part of the inductive current, which will reduce the currents from the central reactor. This report will look at the factors related to grounding systems and how these factors affect the ground fault currents. The purpose of the report is to give recommendations to Umeå Energi on where in their grid they should install additional local reactors and also which factors they should consider when doing future expansions and rebuilds of their power grid.