Presented Without Proof(an exercise for the reader)

My first industrial ecology paper has just been accepted for publication in “Resources, Conservation, and Recycling,” titled “Material Flow Analysis of polyethylene terephthalate in the US, 1996-2007.” Polyethylene terephthalate (PET) is the “#1″ plastic material used to make bottles for soda and bottled water, as well as a steadily increasing stream of clamshell containers. It’s the most-recycled plastic. (update: available online as of May 15, 2010: doi:  10.1016/j.resconrec.2010.03.013).

Here’s the abstract:

We present a material flow analysis (MFA) for polyethylene terephthalate (PET), a synthetic polymer, in the United States for the years 1996-2007.  We model the industrial metabolism of PET as a network of flows linking stocks and processes.  The most common worldwide use of PET is in textile production, but in the US it is increasingly used to make disposable beverage containers for transporting water, carbonated soda, and other beverages. Bottles made from PET are the most-recycled plastic product in the US by mass and by recovery rate, and thus the PET material flow system constitutes an ideal case study of polymer recycling. We find that total consumption of PET resin grew at 2.7% per year over the period of the study, reaching 5.01 million metric tons in 2007. This growth was driven largely by the beverage packaging market, which accounted for 55% of consumption in 2007. About a quarter of PET bottles are collected for recycling, a number that has fluctuated widely but kept pace with consumption.  However, domestic capacity for reclamation of post-consumer PET has not grown as quickly, leading recyclers to export increasing amounts of post-consumer material. Manufacturers have also imported secondary PET in increasing amounts. Reclaimed PET accounted for 6–9% of total resin demand throughout the study. While polymer recycling appears to be viable, efforts to improve material efficiency are confounded by low collection rates and a lack of reclamation infrastructure.

For me, one of the highlights of reading a material flow analysis is the sankey diagram, a flow diagram in which the lines representing flows have widths proportional to the magnitude of the flows.  Creating sankey diagrams has puzzled and intrigued many, because they are both a very good way to convey information and difficult to create in an automated fashion.  I used LaTeX and PS Tricks, of course, to make my sankey diagrams, about which I’ll post more soon.  But for now, here is a sneak preview of the output.  This document is programmatically generated using a data file to indicate the widths of certain primary flows.  The remaining flows are computed arithmetically by TeX.

The hard part, of course, is laying out the figure.  But I’ve convinced myself that the fruits of the hard manual (well, emacs-based) labor are worth the hassle. Click on the figure to download the PDF.