Natural Oil Polythiols and Polyols

1

• Natural Oil Polythiols and Polyols
• A Life Cycle Comparison
• Thomas A. Upshaw, William J. Fisher, Eric J.
  Netemeyer
• Chevron Phillips Chemical Co., LP
• ACS Green Chemistry Engineering Conference
• June 25, 2008

2
Outline

• Study objectives
• Modeling tools and information sources
• Modeled systems and assumptions
• Mercaptanized soybean oil (MSO)
• Petrochemical (flexible polyether) polyol
• Castor oil
• Soy-based polyol
• LCA Methodology
• Impact category results
• Conclusions

3
Soy Polythiol MSO (Polymercaptan 358)
4
Objectives

• Develop a soy polythiol life cycle inventory
  (LCI) platform for product life cycle assessment
  through the product manufacturing stage
  (cradle-to-customer)
• Compare life cycle environmental impacts using
  updated LCI data for vegetable oil and
  petrochemical (polyether) polyols to quantify the
  benefit of using a renewable oil as raw material
• Future assess process changes and new process
  technology for reduced environmental impact

5
Life Cycle Modeling Tools

• SimaPro 7.0 software, using SimaPro 7.0 database
  and U.S. LCI database
• BEES (Building for Environmental and Economic
  Sustainability) impact model
• NIST sponsored EPA supported
• Methodology used by USDA BioPreferred program
• Conducted in accordance with ISO 140401997(E)
  standard
• TRACI (Tool for the Reduction and Assessment of
  Chemical and other Environmental Impacts) EPA
  life cycle impact assessment method

6
Data Sources

• Soybean data
• Agriculture data from U.S. LCI database (NREL)
• Processing data from NREL LCA report on biodiesel
  1998
• Soy Polythiol Chevron Phillips Chemical Co.
• Process inputs estimated from commercial
  production facility, assuming conventional H2S
  process technology
• Soy-based Polyol
• 2004 manufacturer-specific BEES input streams
• Petroleum (flexible polyether) polyol
• U.S. LCI database
• Castor oil
• Purdue University article and various internet
  sources
• Incomplete process data supplemented by analogous
  data on other seed oils in U.S. LCI database

7
LCA System Boundaries
LCI INPUTS
LCI OUTPUTS
Crop oil Feedstocks
Petroleum Feedstocks
Upstream Production of Raw Materials
Agricultural production
Process energy
Raw materials production
Air emissions
Water effluents
Materials production, transport
Vegetable oil production refining
Waste
Process energy
Product Polyol or Polythiol Manufacturing Stage
Air emissions
Water effluents
Materials production, transport
Waste
Transportation to the customer
Energy, materials
Air emissions
8
MSO Polythiol Assumptions

• Commercial process design based on known reaction
  conditions from trial runs at Philtex plant
  (Borger TX)
• UV reactor
• Estimated stoichiometric excess of H2S
• Stripping and recycle of H2S
• Known reaction conditions from lab/pilot work
• Conventional energy sources (nat. gas)

9
Petrochemical Polyol Assumptions

• Consolidated proprietary information for 5 North
  American plants, 2003-5 data
• Polyether polyol, glycerin-initiated, 3500 mol wt
  (on average)
• KOH-catalyzed, solvent, water-washed
• 7.6 to 1 wt ratio PO/EO

10
Castor Oil Assumptions

• Complete data were not available
• Significant uncertainty, need better data
• Analogous LCI data for other seed oils were used
  for some LCI inputs (fertilizer usage, energy)
• Since growth and modernization of castor
  agriculture has been occurring, mechanized
  production and irrigation were assumed for 75 of
  production
• 8200 mile transport from India to U.S. market
  assumed before distribution in the U.S.

11
Soybean Oil Polyol Assumptions

• 2004 manufacturer-specific BEES data
• Produced by simple air oxidation of soybean oil
• No further refinement, purification or
  derivatization
• Soy agricultural model
• Not sure if waste/off-grade is taken into account
• 1000 mile transport to customer

This probably represents the most environmentally
benign vegetable oil polyol process possible a
benchmark for comparison of other renewable
products
12
LCA Methodology

• Life Cycle Inventory quantified listing of
  inflows and outflows per 1000 lbs of product
  (built in SimaPro 7.0)
• Converted to equivalent units per 1000 lbs and
  combined into LCIA impact categories (BEES
  impact model)
• Normalized to unitless dimensions corresponding
  to fraction of total U.S. impact per year per
  capita
• Overall BEES environmental score sum of
  normalized impacts weighted by importance
• 2006 BEES Stakeholder Panel

13
LCA Methodology

• Life Cycle Inventory quantified listing of
  inflows and outflows per 1000 lbs of product
  (built in SimaPro 7.0)
• Converted to equivalent units per 1000 lbs and
  combined into LCIA impact categories (BEES)
• Normalized to unitless dimensions corresponding
  to fraction of total U.S. impact per year per
  capita
• Overall BEES environmental score sum of
  normalized impacts weighted by importance
• 2006 BEES Stakeholder Panel

14
Impact Comparison
(Cradle-to-customer)
15

16

17

18
(No Transcript)
19
Conclusions

• LCA is a valuable tool to help assess
  environmental impact of products and processes at
  a more detailed level.
• more standards and complete, up-to-date publicly
  available data are needed to improve general
  utility and consistency.
• Global warming potential and fossil fuel use of
  MSO and vegetable oil polyols are significantly
  lower than for the petroleum-based polyether
  polyol due to the crop oil raw material source.
• Agricultural practices, oil extraction methods
  and shipping also have a significant impact.
• Future use of renewable energy for MSO production
  would result in a significant reduction in global
  warming potential (GWP) and fossil fuel
  consumption.

20
Conclusions

• Next generation process technology currently
  under development may significantly reduce energy
  consumption, GWP and SOx generation (i.e.,
  criteria air pollutant and acidification
  impacts).
• Castor oil was comparable to MSO overall (BEES),
  but better life cycle input data for castor oil
  is needed
• Castor suffered from the use of the solvent
  extraction process and (probably high) estimated
  water and fertilizer use (vs MSO) and
  eutrophication and smog potential were high vs
  soybean oil polyol.
• A best case soy oil based polyol showed less than
  16 the overall impact relative to a
  petroleum-based polyol
• But best case (simple) process does not
  necessarily give a product with acceptable
  end-use properties

21
Acknowledgements

• American Chemical Society
• Jim Pollack, OmniTech International Ltd.
• Anne Landfield Greig, Four Elements Consulting,
  LLC
• Chevron Phillips Chemical Company, LP