A wholesaler bottler and distributor of both imported and domestic alcoholic drinks purchases half its spirits in barrels, which it then blends and bottles (approximately 60 different types of spirits in up to five different bottle sizes). The most time-consuming activity is the changeover of the bottling line for a different bottle size. This takes one full day. Therefore, products that have the same bottle size are bottled one batch after another, saving the bottle-size changeover time (however, there is a smaller setup time for spirit, bottle, and label changeover). ROP of the one-litre bottles is set to one month of demand because they are usually bottled once a month. Each time, approximately 10 different sprits are run, one after another. The actual bottling time is negligible (only two days for all 10 spirits).

Suppose now is the beginning of June, and there is going to be a changeover to the 1L bottles tomorrow. The demand forecast of this item in June is 312 cases. Assume that the demand has no trend or seasonality. The setup cost per bottle run for vodka is $73.23. The cost per case of vodka is $29.31. The carrying cost rate is 15 percent of unit cost per year.

After talking to the supervisor, it became clear that he does not use the EOQ to determine the batch size. Instead, he produces enough 1L bottles of a spirit so that this quantity plus any on-hand inventory is expected to last until the next production run for 1L bottles, which is one month later. However, for items with uncertain demand such as vodka, he produces a quantity that, together with any on-hand inventory, is expected to meet two months of demand. Assuming that the standard deviation of monthly demand for vodka is 98 cases, what service level is the supervisor implicitly using? Round your answer to 4 decimal places.