Case :

Winery: Developing an Optimal Blending Plan

Daniel Fowler, senior vintner at Napa Winery, had been put in charge of developing an optimal blending plan for the upcoming season. This assignment was the result of a recent Napa Winery board meeting where the CEO had presented her ideas regarding the use of analytics for enhancing profits while at the same time not affecting quality. Industry reports indicated that a growing number of the major wineries were using analytics to assist in the wine-blending process. The board meeting had concluded with the CEO tasking Fowler to develop an analysis and report his findings to the board at next month's meeting.

The United State has become the largest wine market in the world, with sales approaching $40 billion annually. Typically, two types of wines are produced: varietals and blends. Wine blending is the process of combining several grape varieties to achieve a characteristic that is lacking in the original grapes. There are several reasons why a vintner might want to blend wines, including: (1) enhancing aroma; (2) improving the color; (3) raising or lowering the acidity level; (4) raising or lowering alcohol levels. The process of wine blending contains both objective and subjective components. Alcohol level is an example of an objective standard.

Napa Winery was one of the premium wine producers in the state and had recently begun to sell its products on a global basis. The winery produced and distributed a wide range of premium wine, including its flagship - CS Wine. The firm's management was considering employing prescriptive analytics as a means of improving the wine-blending process. Typically, wines were produced from a blend of several types of grapes. In producing these blended wines, the vintner had to take into consideration both grape characteristics and government regulations. Each of the candidate blends was then subject to a series of taste tests. In those cases where the candidate wines were found to be unacceptable by the tasters, a set of new products was often produced. The vintner planned to use prescriptive analytics to help develop an optimal blending strategy and assumed that all bottles produced could be sold. More specifically, the vintner was going to undertake a comparative assessment of Napa Winery' premier CS Wine product sector. The three specific production wines from this sector were:

Vintage CS Wine, which wholesaled for $9 per bottle

Non-vintage CS Wine, which wholesaled for $5.50 per bottle

Non-vintage M Wine, which wholesaled for $2.95 per bottle

Listed below are the winery objectives and government regulations.

Winery objectives and specifications -

Maximize net profit.

The acidity level of CS Wine cannot exceed 0.7 grams per 100 milliliters.

The vintage CS Wine must not contain more than 0.2 per cent sugar.

The non-vintage CS Wine must not contain more than 0.3 per cent sugar.

The acidity level of M Wine cannot exceed 0.3 grams per 100 milliliters.

Government regulations -

All wines labeled varietal (e.g. CS Wine) must contain at least 75% of the named grape type.

All wines must contain at least 10% and no more than 15% alcohol level by volume.

All vintage-dated wines must contain 95% blending grapes from the year on the bottle label.

All vintage-dated wines must also report the viticulture area on the label and must contain at least 85% blending grapes from this area.

Presented in Exhibit 1 are the characteristics of the four blending grades along with available quantities and associated costs.

Exhibit 1.Grape type characteristics, quantities and costs

Grape Type

Viticulture

Vintage

Acidity (gm/100 ml)

Sugar (%)

Alcohol (%)

Quantity (bottles)

Cost ($/bottle)

CS grapes

Zone 1

2011

0.35

0.12

13.5

50,000

2.35

CS grapes

Zone 2

2010

0.75

0.25

15.3

60,000

2.60

CS grapes

Zone 2

2011

0.55

0.30

11.5

30,000

2.10

M grapes

Zone 1

2010

0.25

0.08

15.7

200,000

1.55

Please refer above case, and help with following questions,

Problem statement.

Relevant background and facts.

Summary of key issues/decisions.

Proposed solution methodology. For example, for a linear optimization problem, you should identify the decision variables and key constraints.

Consider the first grape type listed in Exhibit 1 in the case: . Can the winery produce a wine consisting only of these grapes? Can the winery produce other wines that include these grapes and other types of grapes?

Using answer to the question above, identify all of the different grape-wine combinations that are possible.

Formulate a linear optimization model decision variables, objective function, and constraints that company could use to achieve its objective.