Dimensioneringsmetoder för minskad kapitalbindning i säkerhetslager - Utvärdering av dimensioneringsmetoder och förbättring av praktisk användbarhet

Sammanfattning: Stock binds capital and this should be of highest concern for companies to minimize. Therefore it is important to design their reorder points in such way that service levels are met while keeping as low capital as possible. How well do the most common dimensioning methods perform with respect to the capital in safety stock and how the parameter is set in the best-of-breed methods to meet a given service? Objectives The project aims to evaluate different design methods of safety stock in terms of capital tied up in safety stock and make the results practically useful. Methods Stock dimensioning methods performance and outcomes evaluated by using a simulation model, where real data from five companies has been used. Conclusions When comparing different dimensioning methods for reorder points is the result that it differs quite a lot in how much capital is tied up in safety stock and the number of units needed in safety stock to obtain desired service capabilities. The best method from the study, in keeping a high order line service with a minimum tied up capital is the shortage cost per order line. When comparing which parameter value that corresponds to a given order line service the methods differs. The lack of correlation between obtained service level and parameter value is especially apparent for time and cost methods. The only method that is fairly consistent in all simulation cases is fill rate (Serv2) whose parameter value is relatively close to the required service level. Since Serv2 is the method that gives the best match between the parameter value and required service level, it can be used to dimension the other methods more accurate. By using Serv2 based methods opens the possibility of dimensioning safety stock in advance with general safety time or shortage cost methods. The study also shows that the best and most accurate method to dimension security layer to achieve a required level of service, given that the demand follows the same pattern, is simulation of demand history.