BACARDI METHODOLOGY FOR CALCULATION
OF ENERGY & WATER DATA

Bacardi produces a large range of products, and each product is unique and requires its own method of production (e.g., distillation, aging, blending and bottling). These production processes may differ considerably and some are inherently more or less resource intensive than other products. This makes calculating a meaningful figure for our total global energy or water usage much more difficult, particularly for distilleries. To address this difficulty, Bacardi uses an indexing method so that a change in the mix of products that we make does not obscure how our operations are actually performing.

Distilleries typically produce a single type of product using a unique process. Some processes are considerably more resource intensive than others i.e. a “pot still” producing whisky is more energy intensive than a “column still” producing rum and hence even small shifts in the relative quantities of spirit produced can distort actual productivity changes. To alleviate this problem, Bacardi tracks changes in productivity for each product category separately.

Firstly, a measure of the resources required (resource intensity factor) is calculated for each category:



This factor is then normalized to a common value in the base year. Bacardi has chosen to use 100 for base fiscal year 2006 and all product categories are set to the same index value for the base year.

The index for subsequent years is calculated thus:



To calculate the global roll-up, a weighted average index for that year is calculated by multiplying each product’s share of total production:



Using this process, changes in categories of product sold don’t distort overall performance, but as certain product categories grow relative to the rest, they become more important in determining the overall index. The overall index for the target year is calculated by summing the fractional indices of all the product categories.

An index for bottling can also be calculated and used to calculate an overall combined index. The combined index is calculated by taking a weighted average of the two operations (distilling and bottling), but since the production units differ, the weighting is based on the respective amount of resource consumed in the target year. The index eliminates the distortion caused by changes in the relative amount of product bottled vs. the amount distilled.

To demonstrate how the index method corrects for distortions, consider a simple case in which all operations measure no change in efficiency, but the company’s production shifts toward less resource intensive products and does proportionally more of its own bottling.

Table 1 Although all distilleries had no real improvement in efficiency, the shift to less resource intensive products results in an apparent 10% improvement in overall efficiency.



Table 2 Bottling also showed no real improvement in efficiency, but because bottling production increased more than distilling, and bottling is less resource intensive, the company’s overall resource efficiency appears to have improved by 20%.



This simple example shows how our performance can be distorted by changes in products and production activities. All operations were static, i.e., they had no change in efficiency, but the company’s overall “efficiency” improved by 20%. These shifts can be externally driven (e.g., by market forces) or internally driven (e.g., decisions on which products and operations to outsource). To eliminate these distortions and avoid incentives to use outsourcing decisions to drive a flawed metric, we track performance using indices.

Table 3 below shows how performance is aggregated into a single performance indicator. As we would expect, the index shows no change in efficiency (0%) for the distilling roll-up and for the company’s combined index, consistent with the operations’ actual performance.