FIGURE 17.2 Chemical tree of tetrahydrofuran derived from natural sources.
Tetrahydrofuran Butanediol, -1,4 Butyne, -2, diol, -1,4 Acetylene Natural gas Natural gas (unprocessed)
1,000 1,291 1,129 507 763 778
Formaldehyde Methanol Natural gas Natural gas (unprocessed)
1,240 577 356 363
Water (untreated) Water for reaction
246 246
Oxygen from air Air (untreated)
324 324
Hydrogen Naphtha Oil (in ground)
54.2 190 193
Oxygen Air (untreated)
190 263
Oxygen from air Air (untreated)
125 125
Water (untreated) Water for reaction
82.7 82.7
FIGURE 17.3 Chemical tree of tetrahydrofuran derived from a synthetic route.
given chemical or product, there could be several possible processes to evaluate, and the chemical trees are likely to look very different.
When talking about more complex systems, the chemical trees of all the substances or materials that are combined to give a specific product can be quite large and cumbersome. For example, the components required to produce an automobile are on the order of thousands, and almost every one of them requires a separate production process. Figure 17.4 shows a representation of the complexity of the chemical tree for one development route to produce a single fine chemical, which also happens to be the regulatory starting material of a commercial active pharmaceutical ingredient.3As can be seen, the complexity of the chemical tree can expand rather quickly.
There is still another layer of detail and complexity that can be added to the chemical tree and that involves identifying which particular manufacturing plant would be producing the chemical or the part. This is what a supply chain is. In the current economic environment, supply chains very frequently become supply networks, as companies tend to use several suppliers for a given good, and in turn the suppliers may have several suppliers for each part or chemical. These suppliers can be located anywhere in the world, and this usually increases the complexity of the overall supply chain. Figure 17.5 represents the potential pathways for a simple supply chain (network) for a product that requires three parts, A, B, and C. Part A can be procured from either supplier 1 or 2, part B can be procured from supplier 2, 3, or 4, and part C can be procured from supplier 3 or 5. Each parts supplier in turn needs raw materials for manufacture, which can come from one or more suppliers. This is a simple example, but some production systems can be rather complicated.
Example 17.1 In Example 16.5 we saw a comparison among four suppliers of ibuprofen:
two in the United States, one in India, and one in China. Provide an example where the differences in the supply chain affected the results of the ecoefficiency evaluation.
Succinic Anhydride Sodium Borohydride
Methylene Chloride
Sodium Bicarbonate
Product HCl
Hexane
Dichlorobenzene
Carbon Disulfide Benzene
Aluminum Chloride
Thionyl Chloride Toluene
Water NaOH
FIGURE 17.4 One route to producing a fine chemical product used in the manufacture of active pharmaceutical ingredients. Each shaded dot represents a manufacturing or extraction process.
(Source: ref 3.)
Solution In Example 16.5 we saw that the toxicity potential is heavily related to the type of process that is used for the production of ibuprofen. After accounting for those materials having toxicity potentials, we found that the Indian and Chinese syntheses use materials such as dichloroethane, aluminum trichloride, triphenyl phosphine, and other substances of significant toxicity. We also found an increase in the environmental impacts from the waste streams associated with these processes. As more hazardous materials are used, the more difficult it is to control emissions and potential accidents. Consequently, the Indian and Chinese supply chains were found to be less ecoefficient than those of their U.S.-based counterparts.
Additional Points to Ponder How would this be different if the production processes in India and China were contained? How would that affect the price structure?
In terms of measuring, managing, and minimizing the environmental or sustainability footprint of production systems, the more complex the supply chain (network), the more complex it is to estimate the environmental or sustainability footprint of the system. We can add to that complexity the fact that suppliers might change from time to time and that market fluctuations might affect the way that products and processes get to our homes.
One way to obtain the environmental footprint information for a supply chain is to obtain it directly from the first layer of suppliers. Generally, this is an approach that would require more resource and time to complete, and the results would depend on the market share that the buyer has for a given supplier. For example, large retailer chains such as Tesco and Wal-Mart have recently started to measure the environmental impacts related to the packaging used for products sold at their stores. Wal-Mart has, for example, begun
Manufacturing Facility
Supplier 1
Supplier 2
Supplier 3
Supplier 4
Supplier 5 Product
Part A
Part B
Part C
Material V
Material W
Material X
Material Y
Material Z
Supplier 6
Supplier 7
Supplier 8
Supplier 9
Supplier 10 Material
V
Supplier 11
Supplier 12 FIGURE 17.5 Potential pathways for a simple supply chain.
produces a normalized score between 1 and 10 (10, good; 1, bad). This score is a normalized weighted average of the following packaging metrics:
. Greenhouse gas emissions from packaging production
. Sustainable materials content
. Average transportation distance
. Package/product ratio
. Cube utilization (volume efficiency)
. Recycled content
. Recyclability
. Renewable energy to power each facility
. Innovation (different from energy)
The suppliers provide the information to the retailer for each of these nine metrics and a raw score is calculated from weight-based formulas. The raw scores for each metric provide a means of “ranking” one product’s packaging against another product’s packaging for a given segment of products (e.g., beauty products). This ranking is used with products having similar packaging to calculate a final normalized score. It should be noted that the score will vary over time, as it is based on the number and type of packaging data that is obtained from other suppliers.
As you can imagine, this is not a simple task to set up and coordinate. It is not always possible to request and capture information from suppliers within the time frame required for decision making. In this case, another way to estimate the environmental impacts of complex supply chains is to use streamlined life cycle methodologies. As we saw in Chapter 16, a streamlined LCA covers the same aspects as a full LCA, but at a higher level. One can use streamlined life cycle methodologies to get quick estimations of supply chain impacts for a given activity, provided that the groundwork to understand the system boundaries, goal, and scope of the study has been completed. There are also collections of life cycle impact data that may contain generic data for the production of materials and chemicals. These databases can be used directly for a streamlined assessment, followed by some quality analysis that provides an idea of the uncertainty and sensitivity of the numbers.
Example 17.2 Estimate the environmental footprint of the materials needed to produce a fine chemical with a formula weight of 174. The raw materials needed are listed in Table 17.1.
Solution If life cycle inventory data are available from commercially available databases, that would be one way to estimate the footprint of the materials. However, for this example it was not possible to obtain information about the di-N-oxide or the sodium dithionite. In view of that, the environmental life cycle impacts of the materials needed to produce this specialty chemical were estimated using GlaxoSmithKline’s FLASC tool (fast life cycle assessment for synthetic chemistry).4As discussed in Chapter 16, this streamlined LCA tool was designed based on a full LCA, and it is intended for use in making quick comparisons of different synthetic chemical routes. The scores are shown on a scale of 1 to 5 (1, lowest
impact; 5, highest impact), and the user can get a quick, high-level benchmark of the impacts of the materials used in the synthetic route. This assessment is typically performed in less than 30 minutes (see Figure 16.13). In this case, the bill of materials (material name and mass required) and the molecular weight were entered into the tool. For the materials that were not in the database, life cycle environmental profiles were estimated from averaged data.
As can be seen in Table 17.2, the environmental life cycle impacts were estimated for each chemical needed to manufacture the desired substance. These estimations include manufacturing, energy production, waste treatment, and transportation. The numbers provided in each cell represent the estimated contribution of that chemical to the corre- sponding impact. The totals at the end of each column represent the overall estimated environmental impact for the materials in the production of this chemical.
Those totals were then normalized by an internal benchmark and the normalized scores rolled up into a single high-level score using the geometric average of the scores of individual impacts, thus giving the same weight to all impacts. The high-level output is shown in Figure 17.6.
Additional Points to Ponder Which chemicals make the largest contribution to the environmental footprint? What other impacts currently not included would you think of adding to this assessment?