Abstract:
This objective of this master project is to improve a chemical purchasing method
for waste water system. Currently the selected company did not have any appropriate
purchasing system. The chemical order quantities were set by using experiences of
planners. This leads the selected company to face two problems. The first one is a
chemical shortage problem. This occurs since the planner orders less than the
requirement. They therefore have to urgently send this waste water to treat by
outsource suppliers and have to pay for this up to 200,000 baht/month on average. The
second problem is overstock problem. This occurs since the planner order more than
the requirement. For this case, the company has to pay for the obsolete cost of 6,250
baht/month and pay for holding cost of 3,200,000 baht/year. To prevent two problems
above, the author then uses the historical data from January 2005 to December 2007 in
order to forecast the demand in 2008. Three forecasting methods were selected
namely, linear regression, moving average, and weight moving average. All of them
were applied by seasonal index. The result shows that the linear regression with
seasonal index obtains the forecast demand with minimum errors. The author then uses
this forecast demand to calculate the purchasing lot size by using three methods called
basic EOQ, EOQ with penalty cost, and EOQ with service level. The result shows that
the basic EOQ were recommended. By using the recommended method, the total cost
of waste water treatment operations were reduced from 5,339,702 baht to 1,249,741
baht or reduced up to 76.6% compared to the current method.