Greening the supply chain, Part 4

Last of a series

I ended the last supply chain discussion remarking that keeping track of the bits and pieces of a supply chain (resources, impacts, locations, suppliers, transit routes, etc.) is not trivial but that, as we become better at tracking these, the metrics, software tools, analytical methodologies and, even, simple rules of thumb, can be developed and used to get a better picture of where we are and what progress we are making.

I know from a number of advertisements and tech magazine articles that the major players (think SAP, Oracle and ERP and PLM applications - I am sure there are many more) are already well established in the "bean counting" and can provide much of the data needed to get us started at the enterprise level and supply chain at some level.

In an earlier example in this series, we focused on a fictitious auto production  operation with stamping, assembly and sales in different parts of the workd. We saw the sensitivity to energy source impact variations in different locations. Last time we dug a bit deeper into these variations and what they mean in supply chain impact determination.

In this posting, I'd like to present a more practical example based on real data from work done in our lab. And, I'll admit, it's tax time and I have to work on my contribution to the State of California and the Federal Government so this saves me a little time!

This example is based on a paper titled “Development of the Supply Chain Optimization and Planning for the Environment (SCOPE) Tool - Applied to Solar Energy,” which was presented at the IEEE International Symposium on Electronics and the Environment in 2008 in San Francisco. Copies of the full paper are available for download at http://escholarship.org/uc/item/9tq5x8fb.

The paper addresses the need for supply chain optimization and planning with the environment in mind, using reasonable metrics as part of the decision making process. The proposal is for a tool (called SCOPE as defined in the paper title) and the example applies the tool to solar panel manufacturing.

Renewable energy systems are being developed to satisfy three main goals: (1) provide reasonably priced energy (2) mitigate climate change (3) provide energy independence.  The life-cycle environmental impact of energy supply can be reduced through research on materials, product design, manufacturing, and the supply chain; the focus here is on the supply chain because environmental tradeoffs at this level are generally not considered in new energy development.

The supply chain is defined as the set of suppliers required for a complete and successful final product, and the interconnecting network of these suppliers around the globe as detailed in previous postings in this series. The supply chain has been found to impact up to 25% of manufacturing costs in typical products, and preliminary studies indicate that environmental impacts may be similarly distributed. In this study we worked with SolFocus, Inc. (see http://www.solfocus.com/en/) a manufacturer of concentrator photovoltaic systems.

An initial assessment of SolFocus Inc. concentrator photovoltaic systems found transportation to be 10-20% of the lifecycle energy demand when panel transportation to installation site and glass transportation to assembly were included. Additionally, our research has shown that the strategy to minimize greenhouse gas emissions depends both on the electricity mix at the customer and transportation distances.

To do this correctly, environmental supply chain considerations can and should be incorporated early in the design process to ensure the greatest possible reduction in impact. Previous solar energy assessments, while thorough in their execution, have not focused on the climate change mitigation potential of a re-organized supply chain or installation location variables.

The SCOPE tool is a hybrid LCA tool and incorporates the following:

1. Electricity mix and resource differences throughout the supply chain since parts may originate from all over the globe, such as China, India, the U.S.A, and elsewhere. This approach parallels today's economic assessment for production, where manufacturing location decisions are influenced by economic decisions such as labor costs, energy costs, local regulations, resource availability, flexibility, and lead times. (But, the tool does not include these latter elements - yet.)

2. Transportation emissions and energy demand.

3. Electricity distribution and circularity.  When determining the electricity that is offset by a new solar installation the "circularity" or distribution losses of electricity supply are usually not considered. Demand for electricity requires extra production to account for electricity that is lost in transport to consumers (distribution losses) or internally demanded by the energy sector (circularity).

The proposed SCOPE tool's basic architecture and underlying hybrid LCA methodology are presented in the paper and energy and greenhouse gas metrics are used by the tool to assess alternatives so check the paper for details. A case study of preliminary results for SolFocus Inc. is presented to establish the feasibility, applicability, and usefulness of SCOPE.

The metrics used are:

- energy payback time (EPBT) which indicates the number of years a technology must produce electricity, thus offsetting the use of primary energy from another electricity source, to offset the total energy required over its lifetime (including manufacturing, transport, installation, etc.)

- energy return on investment (EROI) calculated as the lifetime of the product divided by the EPBT  EROI indicates how many MJ of primary energy are saved from consumption for every MJ of primary energy consumed.

- greenhouse gas payback time (GPBT), an analog of energy payback time for green house gas emission, and

- greenhouse gas return on investment (GROI). Similar to  EROI, GROI indicates the GHG emissions prevented for every unit of GHG emitted encouraging the fastest route to reducing energy related greenhouse gas emissions.

To illustrate this approach an example is presented. In this collaborative project, we looked at a particular  utility scale concentrator photovoltaic system SolFocus was developing. It is comprised of a solar panel, a "tracker" support structure that moves the panel to maintain incident solar exposure during the day, controller in the tracker, etc. Although the design and manufacture were still under development,  available cost estimates and preliminary manufacturing data were available. A mockup of choices for producing a solar panel is shown in the figure below. The double boxes indicate one possible supply chain.

To illustrate the variability in energy and GHG metrics, four scenarios are considered: (1) no transportation over the life of the supply chain (2) transportation of goods across the SolFocus supply chain using the most efficient methods possible (truck, rail, water freight) (3) transportation using only air freight as a worst case scenario (4) same as 3 except installation in France rather than Phoenix with a DNI of 5.3 kWh/m2/day. DNI refers to "direct normal irradiance" and gives a measure of the direct solar energy available at a location. Not surprisingly, locations in sunnier locations (as in closer to the equator) are better for solar energy. Phoenix is considered a good site while France is considered marginal; DNI values can be even higher in Africa, Australia and other parts of the southwest United States. If you want to see what the DNI is for your location (in the US) see http://www.solarpanelsplus.com/solar-calculator/, for example.

For this assessment, installation is assumed to occur in Arizona, USA with a DNI of 6.9 kWh/m2/day. The panels are assembled in India, and most components come from China, India, Spain, or the U.S. The installation is utility scale and assumed to replace rather than supplement the local electricity mix; therefore production, circularity, and distribution of the current electricity mix are offset.

Although the SCOPE tool is still under development in LMAS, the application of the SCOPE methodology to this particular solar system, given available data, results in the supply chain tree shown in the figure below. Note that in the figure transportation is not yet included for every component.

The results of the analysis are shown in the table below and indicate the influence of transportation

in energy and GHG metrics. In reading the table, smaller values of payback time and impact of GHG/kWh are better! The GPBT shows the largest sensitivity to installation variations because it is directly proportional to the conversion factor for GHG per unit of electricity (recall last posting) and the circularity. Each metric is also sensitive to the type of transit used throughout the life-cycle as seen between scenarios 2 and 3. These results indicate the potential for supply chain and installation optimization using the SCOPE tool. The impact of variations in electricity mix at each supplier site on these metrics have not yet been explored.

Preliminary results for SolFocus concentrator systems across the range of scenarios tested indicate that the EPBT of SolFocus Panels can vary from 0.6 to 5 years depending on installation and manufacturing locations. The GPBT can vary from 1.1 to 49 years depending on the same factors, and is seen to be more greatly sensitive to location factors than an energy-based metric.

But, there is much to be done to complete such tools - specially for alternate energy applications. Energy use and greenhouse gas emission metrics are discussed here because energy use is relevant to the efficiency of solar technology and greenhouse gas emissions are relevant to climate change; however an additional key concern of climate change is water scarcity. Solar has the distinct advantage of not requiring water during its use-phase; therefore, the installation of solar to replace thermal power plants (which consume nontrivial amounts of water per kWh generated) in water scarce regions of the world could prevent water use and thermal pollution of waterways. Tradeoffs then emerge between EROI, GROI, and water scarcity that will require further investigation and understanding to design minimal impact manufacturing supply chains. Also, additional environmental metrics could be added as well, such as toxicity and acidification potential.

The tool does not currently account for travel by engineers, managers, and executives to work collaboratively, ensure quality control, and provide feedback. Also, estimation of error will be an important final step to an analysis using SCOPE. As inputs on costs, weights, distances, and more are entered into the tool, confidence intervals could be included that would result in a confidence interval on the final solution.

Finally, it is important to note that while SCOPE provides decision makers the ability to understand environmental tradeoffs between supplier location and transportation, decision makers must also consider lead times, flexibility, and quality of suppliers before making a decision; cost and operations considerations must eventually be included in SCOPE for it to be a viable and useful tool for decision makers (or, conversely, the tool must be integrated with existing software for estimating these elements as part of an integrated package.)

Although this example is more complex than the automotive example, there are obviously, still many layers to be considered. But, in this realistic example, the potential for assessing and optimizing (or at least choosing among several alternatives) the manufacturing supply chain for a real product is illustrated.

I hope you found this series interesting and that it gave some insight into green supply chains from a manufacturing perspective. For sure there will be more reference to green manufacturing supply chains in the future. It is one of the critical elements in green manufacturing.

Now ... back to my taxes!