We are saddened to report that Professor Dornfeld passed away in March, 2016. If you enjoyed his blog, please consider making a contribution to either of two funds at UC-Berkeley that have been established in his memory.

David A. Dornfeld Graduate Fellowship
David A. Dornfeld Scholarship

Friday, March 26, 2010

Greening the Manufacturing Supply Chain, Part 2

Part of a series

We started a discussion about supply chains with respect to environmental impacts related to the various actors in the chain. We'll continue here with an example.

But, first, a story from the trenches.

I've mentioned Interface Carpets before as an example of a company dedicated to green manufacturing and on a path to sustainable business practice. An AutomationWorld article recently on "Sustainability Leads To Next-generation Manufacturing" (by Gary Mintchell, March 10th, 2010; download the article at http://www.automationworld.com/print.php?id=6671) gives some statistics on how Interface has done in the words of its founder and retired CEO Ray Anderson. Anderson talks about the 12 year process starting from the Kyoto Protocol in 1997 which he says "was widely derided by his fellow CEOs that sought to reduce greenhouse gas (GHG) emissions by about 7 percent in the United States by 2010. Others were afraid that meeting that goal would drive them out of business."

His statistics show the opposite is quite true if pursued consistently and with metrics to measure where you are and where you are going. Anderson states "Interface’s performance by 2008 revealed a reduction of 71 percent in absolute tons of GHG emissions while sales increased by two-thirds and earnings doubled. Interface consumption of fossil fuels per square yard of carpet was down 60 percent, and waste reduction measures put a cumulative $405 million of avoided costs directly into the bottom line."

Energy per unit of product reduced by 60%! Recall the "master equation"? It defined impact as a product of three terms: population, per capita GDP, and impact per unit of GDP. I noted that the only "knob" we can twist on this equation is impact/GDP. And 60% reduction as realized by Interface is on track to meet the reductions needed that we spoke about several postings ago.

Anderson goes on to enumerate the tangible benefits of following this path - strong competitive position, growth in sales, higher profits (since costs are down) and adding to employee satisfaction being part of such an organization.

Back to supply chains (although if you are a builder of commercial space then Interface is probably in your supply chain!). A recent article in Environmental Leader (Feb 1, 2010, see http://www.environmentalleader.com/2010/02/01/survey-56-of-cdp-members-may-cut-out-suppliers-who-dont-manage-carbon/ for the full article) discussed a survey of the Carbon Disclosure Project (CDP) regarding their attitudes towards suppliers (supply chain) that do not manage their carbon. These are real companies in the survey (like PepsiCo, Dell, Google, IBM, Kellogg, HP and Unilever.)

The results are enlightening. Over 1000 suppliers to these companies were surveyed. Survey reported that 38% of the supply chain respondents have some type of carbon reduction targets in place. Of these respondents, almost two thirds report Scope 1 and Scope 2 emission. Scope 3 emissions are reported by 8%. Strikingly, 56% of the CDP members (remember the big companies listed above?) say that they may eliminate suppliers who don't manage carbon.

You can get the full CDP report (download the pdf) online at https://www.cdproject.net/reports.asp.

Now, let's look at an example of the impact of supply chains and, in particular, energy mix and transportation.

This example comes from one of my PhD students, Corinne Reich-Weiser, who is studying decision-making methodologies as applied to reducing the greenhouse gas emissions of manufacturing.

In this example,  the effect of minimizing global emissions on a supply chain is considered by looking at a simple vehicle manufacturing global supply chain with four candidate facility locations for assembly and stamping. We assume that the company serves the US market and wants to decide (1) where it should open facilities and (2) how to ship from the stamping facility to the assembly facility, and from the assembly facility to the market. The manufacturing technology, and the product produced, is the same regardless of location.

That is, the energy consumption for stamping and assembly is the same at each facility. But, the CO2 emission will be different at each facility due to the different energy mixes associated with the electricity supplier in that location (recall our discussion about conversion factors for kilowatt-hours to carbon dioxide for different regions of the world and US - this was in one of the earliest postings, see http://green-manufacturing.blogspot.com/2009/07/why-green-manufacturing-part-2.html.) There is only one supplier in this example (the stamping facility), one component (the stamped metal sheets),
and one manufacturer (the assembly facility.) We assume that the rest of the components needed for vehicle assembly are produced in local facilities. Also, the location and transportation mode are determined independently.

For this example, we assume that a typical vehicle weighs about 1,500kg and the total stamped sheet metal weighs about 1,000kg. The cost for stamping and assembly of one vehicle are $700 and $100, respectively.

The table below shows the estimated CO2 emission levels associated with vehicle stamping and assembly the different areas we are manufacturing in this example derived from EIO-LCA data.


You can see from the table (and apologies for the clarity of the table and figure below) the candidate locations for our stamping and assembly and, eventually, market example.

If a facility is outside the US, then components have to be sent to a port and then shipped internationally. We assume that air transportation is not an option here. There is only one way to ship products internationally, but one can choose either truck or train to deliver the product domestically. Further, emission characteristics will differ depending on the type of vehicle used. Because the carbon emission factor of trains (0.022 kg-CO2/tonne-km) is less than trucks (0.033 kg-CO2/tonne-km) according to Mike Ashby's data, it is always optimal to select trains over trucks if we only consider carbon emission.

One must include some production costs in the analysis as these will vary substantially with location. We can estimate the variable production cost in different areas by considering labor cost, utility cost, and facility rental costs in different countries.

The summary carbon footprint using three different distribution options (minimal economic cost, local manufacturing, and optimal carbon emission) is illustrated in the figure below.


The figure shows the contribution of transportation (hatched), stamping (black) and assembly (white) for each of the scenarios.

The lowest cost option is to stamp and assemble in China and then ship to the US. The minimal cost option  emits more than twice as much CO2 as the minimal CO2 option. The minimum carbon emission is achieved when the stamping is done in Germany with  assembly in the US. This is because the energy mix in Germany is half the impact of that of the US. The transportation emission is also smaller compared to other countries except the US because of the relatively short distance between our assumed location for stamping (Stuttgart) and the port. Because local manufacturing (i.e. in the US) saves on dramatically on transportation costs and emissions, it ends up in the middle in this example (due to energy mix issues again.)

This example shows that supply chain designs change when environmental impact is considered. This was arguably a very simple example with only a few options for organizing the supply chain. But, as seen in the lower figure comparing the results, there is a tremendous difference in these three scenarios. The optimal chain is less than one half as impactful as the "minimum cost" chain.

So, specially if you are a supplier to any of the CDP companies, it might make sense to look at all your supply chain options!

Saturday, March 20, 2010

Greening the Manufacturing Supply Chain


First in a series

The last couple of postings covered some examples of the dramatic changes in production and materials processing that are going to be required to meet the goals for green house gas reduction set by various entities. These will go well beyond efficiency "tweaks" and involve substantial new technology wedges. We reviewed a couple of examples. We'll get back to this again I'm sure.

If you Google "green supply chain" you get over 12 million results. If you search the same term on Environmental Leader you get quite a list of related articles. This is now one of the hot topics and, as with most of things green, there is a wide variety of interpretation of what, exactly, the term means.

So, here is my interpretation!

First of all, I am not a supply chain expert. I have come to realize over the past few years that there are companies that make almost everything themselves (fewer and fewer but they still exist) and rely on some outside suppliers and there are companies that make almost nothing themselves and rely on an extensive network of suppliers. And, most important, I've come to know some supply chain experts here at Berkeley (particularly Professor Max Shen in Industrial Engineering and Operations Research) so I am emboldened to charge ahead!

Let's start with some definitions. Remember the Ricoh Comet Circle? (Use the search function on the blog page to find this if you need a refresh!) That shows the path of a product to the consumer and the fate of the product after it leaves the consumer. The consumer can be you or me, a company using a machine (product) or consumable, etc. This diagram visualizes the forward and reverse supply chain.

A supply chain can be defined as the "network of retailers, distributors, transporters, storage facilities and suppliers that participate in the sale, delivery and production of a particular product" (source: http://www.investorwords.com/4823/supply_chain.html). These are usually stratified into first tier, second tier and, sometimes, third tier suppliers depending on where they are in the "food chain" so to speak.

The image below, from otl.curtin.edu.au/tlf/tlf2001/ee.html, shows the inter-relationship among these suppliers and manufacturing, etc. including material and information flows.


And, one could add 3rd tier suppliers to this as well. All the components of the Comet Circle are represented. The locations of these elements in the figure can be anywhere in the world (and usually are) that the company feels the best value can be obtained. The time, cost, and now, environmental impact of all these flows is of major importance.

Remember the discussion of the UK's goal of CO2 reduction and the comparison of the actual vs "apparent" reduction we spoke of a few postings ago? (See http://green-manufacturing.blogspot.com/2010/03/digging-deeper.html for details.) The actual, if you included "outsourced CO2," was moving in the opposite direction of the target reduction. The figure above tells you how that happens - make sure the heavy CO2 contributing elements are outside the UK. Bad for the accounting and, incidentally, for domestic manufacturing!

These supply chains can be impressively complicated for sophisticated products, for example the laptop computer I'm writing this on. This is due to the large number of different parts and elements, the large number of suppliers (at all tiers) of these parts and, importantly, their location and the transport means for getting all this together to result in my laptop.

Our concern here is with these material flows and the embodied energy, water, resources, for all aspects of the manufacture, including transportation and storage/distribution associated with those.

Depending on how much a company relies on its supply chain will determine to what extent that company can affect the impact of its product. Wal-Mart's sustainability initiative is a major example of this new trend. It is estimated that, for Wal-Mart, as much as 90% of a product's associated carbon emissions (transportation, manufacturing, farming, etc.) are from the suppliers. So for a company like Wal-Mart, if they want to reducing their emissions overall, they need to find a way to affect their supply chain.

Recall that a  firm's environmental performance is usually evaluated in terms of energy consumption and carbon footprint. Most companies focus on direct emission (Scope 1) and indirect emission from purchased energy (Scope 2) (recall our discussion of these scopes, see the December 25th posting at http://green-manufacturing.blogspot.com/2009/12/green-new-year.html.) Direct emission from company owned or controlled activities and indirect emission from purchased energy make up only a small percentage of the total supply chain emission, excluding emission from the use phase.

The supply chain emissions will vary with the product and industry. For example, the figure below, derived from the CEDA database from Professor Sangwon Suh at the University of Minnesota, shows the carbon emission in the supply chains in several selected manufacturing industries.


The range is impressive. For some electronics only about 10% of the carbon emission is direct. The bulk of the emissions for computers, more than 80% in the figure, is from embodied emissions of the purchased parts. Ditto for motor vehicles. The lowest supply chain impact in this example, plastic materials and resins, is still showing that less than half of the carbon emission is due to direct emissions and electricity consumption - rest from the supply chain.

Some companies have started to look at overall supply chain carbon emissions. However, they often tend to ignore the interaction between different elements of their supply chains. This could be due to the complexity of tracking material flow through all the elements. These elements include all the components seen in the first figure above.

Ignoring interactions can have deleterious effects. For example, if one changes the shipping mode to a lower-carbon option such as rail, the carbon emission per product will decrease. However, delivery by train may result in a longer lead-time and, thus, necessitate a higher safety stock at the retailer or production facility. In turn, greater inventory at the retailer or plant will increase the energy consumption and carbon emissions of their storage facility and warehouses. At least it will likely require more floor space and the related impacts of that in energy, etc. Thus, the overall CO2 reduction may be smaller than expected or, perhaps, non-existent.

As another example, suppose there is a manufacturing process changeover that consumes more materials and resources. One may want to increase the batch size to reduce carbon emissions. However, larger batch sizes will require an increase in the system WIP, and thus increase inventory level carbon emissions.

We discussed some of the means to evaluate such trade-offs in an earlier posting on lean and green manufacturing since many of these concerns are also central to lean manufacturing analyses.

We'll continue this discussion with a more detailed example next time to show how we account for some of these interactions and weigh the impact of a particular supply chain.

Friday, March 12, 2010

Not Business as Usual


The challenge of keeping ahead of the curve on the reduction of consumption or impact was made clear in the last posting. We saw the expectations of the EU to achieve a 60% absolute cut in yearly carbon emissions by 2050 compared to 1990 levels. And we saw how they were doing.

The problem is that consumption keeps increasing (and impact with it) while we are trying to reduce this impact. Thus,  we need to accommodate both reduction in per unit impact (CO2 for example) as well as the increased production with increased demand. One of the key strategies to this is to fundamentally rethink how we process, and re-process materials. The example from Allwood cited last time showed the extent to which this needs to be done for steel. We need to be able to facilitate the loops closest to the consumer in the Ricoh comet cycle to make this work.

This will require a number of substantial technology wedges to pull off.

So, what are some examples? I'm going to start with a couple of examples given by Professor Allwood in the presentation slides I referred to last time. These deal with photocopy paper and aluminum.

Allwood compares existing methods of paper recycling from the office copy paper use with a new process. Traditional recycling collects paper from the user (large and small offices or homes), pulps the paper and adds chemicals to de-ink it (remove the ink from the fiber matrix usually in a foam or froth) and then insert the de-inked fiber back into the papermaking process. The result is new paper with some percentage of recycled paper.

An alternative process uses a novel toner removal technique right in the office. Think of a "reverse copy machine" that takes in used paper (paper with fused toner material on it) from the office and with some type of adhesive, solvent, abrasive,  or laser process (the latter two accompanied by a vacuum tube) removes the toner to yield paper for re-use. A paper by Allwood on these approaches titled, appropriately "Meeting the 2050 carbon target for paper by print removal," (the web link is too long - "google" Allwood and the paper title if you want to see the details!).

Another idea put forward in the Allwood presentation is on recycling aluminum by "cold bonding." Cold bonding is a process where by ductile materials, such as aluminum, are fused into homogeneous masses by pressure as in an extrusion or pinch rolling process. The traditional means for recycling aluminum is to collect, separate, sort and clean the metal from a variety of sources. The material is then melted and cast in a manner close to original production. The ingots of aluminum resulting are then converted into products much as before. This recycling method uses about 5% of the energy needed for production of virgin aluminum (not bad actually.)

But, the cold bonding process would be done on a smaller, more local, process with cleaned aluminum scrap. Deformation under high compression and extension yields a lower strength product but one which is produced in one stage from scrap to product with, according to Allwood, only 1% of the energy of the recycling methodology in the traditional means. So ... now you are down to 1% of 5% of virgin material production energy.

The two figures below show, on top, methods of compression and extension suitable for cold bonding and, on the bottom, a photo from Allwood of the resulting material (aluminum) stock.



Clearly, in these two examples one might question whether or not the resulting material is suitable for all applications. It won't be in some cases. But it will in a lot of cases and offer a substantial reduction in impact while still contributing to demand. And, specially in the paper recycling example, you'd need to consider the materials and impact embedded in the hardware for toner removal.

The last example is much closer to implementation (in fact is available today) and is from a company in San Francisco called Industrial Origami (see http://www.industrialorigami.com/home.cfm ). I ran into this company some years ago when they were getting started and requested some assistance in developing some applications of their novel, and patented, technology.

Industrial Origami (or IOI) has a technology of precision material folding "based on the creation of fold defining geometries which, when put into sheet metal, enable structure and innovative shapes never before possible with traditional technologies. These features called "smiles", control the folding and are responsible for the accurate folding properties." The smiles guarantee certain folded strength and precision of the edge. The pre-cut sheets (both shape and integral "smiles") essentially code the DNA of the final part. A sequence of folding yields a complex three dimensional box for appliances, electronics, automotive components, towers, structures, etc. And the folding process insures accurate dimensions without tooling or fixturing.

There are a number of "green" advantages to this. Complex 3-d shapes can be fabricated in the "flat" and shipped efficiently for local product fabrication. Minimum tooling to produce the final part means low embedded energy and resources in fabrication. Recycling and recovery is enhanced due to minimum fasteners and other attachments. And, in the "Chevy hood to Chevy hood" mode I mentioned last blog - you literally end up with a flat sheet of metal when the product is "deconstructed." It could, within some reason, go back into another product.

One of the applications IOI has been working on is low cost commercial production of light weight vehicle chasses using "using less material, simplified assembly processes, significantly reduced capital investment in a much shorter product development cycle." The image below from IOI's website (http://www.industrialorigami.com/solutions/transportation.cfm ) is an example.


The autobody has additional fasteners (spot welds, for example, to improve crash worthiness) but is basically a "folded" structure. This technology is right in line with the other examples offering "leap frog" advances in material utilization and production efficiency and leading to close to the consumer recovery of materials in line with the Comet circle.

This is getting interesting!

Next time we start talking about greening the supply chain.

Thursday, March 4, 2010

Digging Deeper


The ongoing low hanging fruit discussion here was the preamble to the last posting about scratching the surface (or, as one of my grinding expert friends says ... when it comes to abrasive process research we've only just scratched the surface! Sorry... engineering humor).

It is necessary to understand the magnitude of the challenge and where to best concentrate our efforts. And, sometimes the easiest stuff doesn't do much to advance the cause, specially with respect to greening manufacturing.

Last time I presented some perspectives on what it really means to be sustainable and this referred back to the earlier discussions about "technology wedges." An analysis of some of the potential wedges was presented from the Vattenfall report. Many of those technology wedges were related to manufacturing as noted.

One can find similar data on what kinds of reductions in CO2 emissions (in terms of parts per million in the atmosphere) are needed to get the atmospheric concentration of CO2 at a "sustainable" level. Again, I point out that not everyone agrees with the data and I am not promoting any specific interpretation. But (and a significant but) agencies, states, countries and regions are making regulations based on these discussions, consumers are making choices based on products and companies that respond to this data and companies are changing business plans and strategies based on this. So, in keeping with the "Everett and Jones" philosophy... we'd better be at least watching this carefully (and if you don't recall what this is - search Everett and Jones on this blog page!)  If you check the pages of Environmental Leader (link at bottom) on any day you see a growing list of reports of companies responding.

Ok...we've got the motivation. Now, what will really work.

Last July in this blog referred to a presentation in 2005 by Professor Julian Allwood of Cambridge University on "What is Sustainable Manufacturing" as a good place to start this discussion (see http://green-manufacturing.blogspot.com/2009/08/dimensions-and-metrics-of-green.html for the blog and http://www.ifm.eng.cam.ac.uk/sustainability/seminar/documents/050216lo.pdf for a download of Allwood's slides).

Allwood discusses in detail strategies for reducing the carbon footprint and other impacts of manufacturing. He specially discusses these with reference to targets for reduction set by governmental agencies in the UK and elsewhere. For example, the figure below shows the reduction targets set by the UK and EU to allow surface temperature stabilization. The target is a 60% absolute cut in yearly carbon emissions by 2050 compared to 1990 levels. What is seen is the slope of reductions (in CO2 equivalent)


needed to meet this ambitious goal (in blue), the actual reductions observed over the first few years (and reference to the Kyoto target, in red) and, gulp, these actual reductions adjusted to "off shore" effects. That is, moving the production and associated CO2 generation out of the region of calculation (i.e. out of the UK and EU) in orange. That curve is moving in the wrong direction.

This is not a pretty picture and emphasizes the complexity and difficulty of the task. Smoke and mirrors are not going to get this done.

Since consumption increases annually with additional production of products to meet growing demand fueled by, at least, growing populations (and compounded by increasing expectations of quality of life - recall the "impact equation" we discussed some postings ago on the drivers of impact including impact/GDP) we need to accommodate both reduction in per unit impact (CO2 here) as well as the increased production with demand. A double whammy.

Allwood puts some numbers on this. He summarizes data from an EU project on reducing CO2 in steel production. If demand for steel doubles then stablizes, and every efficiency known is perfectly implemented, the carbon target requires that two thirds of all steel is re-used without re-smelting, and the energy of all forming processes is halved.

Think about this (and keep in mind the Ricoh comet cycle; see http://green-manufacturing.blogspot.com/2009/09/sustainability-angst.html). Two-thirds of all steel re-used without re-smelting! That cuts out a major part of the present strategy for recovery and recycling steel (and many other materials as well.) That means, effectively, if we are to meet this aggressive target (but the type of target many feel is absolutely needed and being discussed by other countries and regions in the world) we'll be taking the steel hood off of our Chevy and re-assembling it onto another with little additional processing! That is, we have to be able to facilitate the loops closest to the consumer to make this work.

Hopefully, this will spur research on and development of a whole host of imaginative re-processing technologies that can cut out the "dirty" part of recycling.

We'll look at some ideas about this next time.

Finally, Professor Allwood has just written a paper to be published in  Environmental Science and Technology Journal (ES&T) detailing some potential next steps. The article, titled "Options for Achieving a 50% cut in Industrial Carbon Emissions by 2050" introduces the idea of material efficiency with reduced primary production. We will delve into this more in the future.