LL1-11 digest — Ch11 MTBE in petrol as a substitute for lead#

Late lessons from early warnings (EEA, 2001), pp. 110–125. Authors: Martin Krayer von Krauss and Poul Harremoës. Harremoës chaired the report’s editorial team. The chapter has no panels. It was written while the case was still open: editing finished 10 May 2001, and Table 11.1 ends “The debate goes on” (p. 121).

Core story#

• Why MTBE was chosen. It replaced lead as an anti-knock agent because it was cheap, made at the refinery, blended easily and could move through existing pipelines (p. 110). Environmental fate does not appear among these listed advantages, which come from a 1996 source, not from decision records. The authors conclude, with a hedge: “It seems that the possibility that MTBE could pose a threat to groundwater reservoirs was never considered” (p. 117).
• How it scaled. European production began in 1973 and US production in 1979. US use “greatly increased” after the 1990 Clean Air Act oxygenate requirement (pp. 110–111). By 1995 it was the third most produced organic chemical in the US (p. 110).
• Weak benefits. Reformulated gasoline as a whole cut emissions, but the air-quality benefits specific to MTBE were small and shrinking. This rests on NRC findings, and on benefits applying mostly to older vehicles that are being replaced (p. 111).
• Harm to water. MTBE is highly soluble, moves with groundwater, degrades slowly underground, and can be tasted and smelled at 5–180 µg/l (pp. 111–112).
• A 1996 US Geological Survey report found it the second most frequently detected VOC in shallow urban groundwater (p. 111).
• Santa Monica lost 71% of its local supply and now buys in about half its water at USD 3.5m/yr (p. 114).
• Unresolved health questions. Cancer evidence stayed borderline (Box 11.1, p. 113). The authors describe “near ignorance” on asthma and endocrine effects (p. 120).

Key evidence#

• Ignored warnings. The 1990 lab-scale warnings (Barker; Jensen and Arvin) drew no regulatory reaction until the 1996 field results (p. 115). The Danish EPA later admitted it dismissed a 1990 warning because tank leaks were “minor” and petrol components were “rarely found in groundwater at the time” (p. 114). MTBE was still not routinely monitored in EU groundwater in 2001 (p. 112).
• Grandfathering. EU rules requiring pre-market notification of new substances (from 1979), which already treated persistence as undesirable, did not cover MTBE because it predated the 1981 cut-off. Its risk assessment reached draft only in 2000–01 (p. 116).
• No “comprehensive” pre-market cancer testing (p. 112). When testing did come, the same data led to divergent verdicts (p. 113):
• IARC: “not classifiable”;
• US National Toxicology Program board: a 6–5 vote against listing;
• US EPA and the White House science council: “carcinogenic potential”;
• EU rapporteur: “borderline” between non-classification and Category 3;
• European Chemicals Bureau working group (November 2000): rejected classification. The authors surmise that threshold reasoning was the “probable basis”; the ECB’s own reasons are not reported.
• The key critique against classification (Dekant) is an expert review obtained by personal communication, titled as an argument against Category 3. The main source cited for “probable human carcinogen” (Mehlman) is, by its title, a short polemical commentary on the NTP vote. Both poles rest on position pieces (pp. 113, 122, 124).
• California cost–benefit analysis: air benefits of USD 14–78m/yr, “essentially the same” for MTBE, ethanol and non-oxygenated petrol, against MTBE water treatment costs of USD 340–1,480m/yr (assuming carbon treatment to below 5 µg/l; “there may be cheaper methods”) (p. 119).
• Responses:
• Denmark dismissed the 1990 warning, set a tentative 30 µg/l limit and an action plan in 1998, and listed MTBE as undesirable in 2000 (pp. 114–115);
• a California Executive Order (March 1999) called for a timetable to remove MTBE no later than end-2002 (p. 114);
• US EPA announced steps to “significantly reduce or eliminate” MTBE in 2000, citing it as a “possible carcinogen” and a threat to drinking water (pp. 114–115);
• EU reports leaned toward tank-focused risk reduction, which the authors characterise, hedging, as treating leakage as “a technical problem” (p. 115);
• US data show even double-walled tanks with detectors “may leak undetected due to improper installation”, which is why the EU’s own consultant report stresses enforcement and monitoring (p. 115).

Authors’ lessons#

• Derived from the evidence:
• Foresight “could have” predicted that persistence would come to be judged problematic (pp. 117, 120). The underlying claim, that competent chemists could have anticipated MTBE’s persistence from 1950s–60s knowledge of ethers, is conceded to be undocumented: “Documentation for this argument has however not been found” (p. 115).
• Persistence, mobility and high volume should have triggered investigation, which might have “acquitted the chemical or confirmed suspicions” (p. 120).
• Industry and regulators were both “short-sighted” (p. 119).
• Recommendations:
• Thorough investigation before any large-volume release of a persistent chemical, reopened as new kinds of harm are recognised.
• Seek alternatives “whenever possible”.
• Pursue risk reduction and R&D on alternatives (p. 120).
• Accept a response costly in proportion to the potential harm (p. 119).
• Advocacy beyond the evidence: radically different transport options (p. 118).

Main mechanisms#

• Judged as a lead replacement, with its own threats “not adequately characterised” (p. 117). How the original comparison was made is not documented.
• Substitution chains: lead to MTBE to ethanol or other ethers (pp. 111, 114–115, 118).
• Regulatory spillover: an air mandate harmed water, and assessment split by medium missed it (p. 114).
• Grandfathering (p. 116).
• Warnings discounted: regulators waited for field-scale proof (p. 115). The Danish EPA cited the rarity of petrol components in groundwater while MTBE went unmonitored; reading that as “unmonitored taken as absent” is our inference (pp. 112, 114).
• Defaults settled borderline evidence (p. 113).
• Costs diffused onto the public (p. 119).
• Opposite time profiles: benefits shrank while harms persisted (pp. 111, 117).

Transferable insights (with strength)#

1. Evaluate substitutes on their own terms, not only against the incumbent (pp. 110, 114, 117). Moderate–strong
2. Regimes that cover only new entrants leave established high-volume items unexamined (p. 116). Strong on the legal fact; moderate that coverage would have changed the outcome.
3. Deployment scale should trigger scrutiny (pp. 110, 114, 116–117). Strong (this case)
4. Hazard emerges from property combinations interacting with infrastructure (pp. 111–112, 115, 117). Strong
5. Loss of use of a shared resource is serious harm even without toxicity (pp. 112, 114, 119). Strong
6. Programmes aimed at one objective can harm another, and fragmented institutions miss this (pp. 110–111, 114). Moderate–strong
7. Test component-specific benefit before mandating a component (p. 111). Moderate
8. Small-scale warnings are discounted until harm is large, and missing monitoring masks problems (pp. 112, 114–115). Moderate
9. Borderline evidence is decided by institutional defaults (p. 113). Strong (descriptive)
10. Testing after deployment leaves uncertainty that outlives the decision to scale (pp. 112–113, 120). Strong on the timeline; moderate that earlier testing would have changed the decision.
11. Irreversibility makes the costs of error asymmetric (pp. 117, 119–120). Moderate
12. Containment depends on implementation and enforcement (p. 115). Moderate
13. Mandated or permitted activities diffuse liability onto the public (p. 119). Moderate
14. Mature alternatives before they are needed (pp. 118, 120). Moderate

Main caveats#

• Hindsight. The foreseeability claim is undocumented (p. 115). The early-1980s list of concerns includes carcinogenicity, which only emerged in the 1990s (p. 117). Early knowledge of taste and odour rests on one unreferenced row of Table 11.1 (p. 121). The “introduction” date drifts between 1973 and the early 1980s.
• Missing decision records. There is no primary evidence on how industry chose MTBE or why the US mandated oxygenates. The “substitute for lead” framing underplays the US oxygenate mandate, which our reading suggests drove most of the documented harm.
• Overstatement. “Everlasting” risk (p. 120) outruns the chapter’s own, mixed, degradation evidence. The authors call reported childhood asthma rates “alarmingly high” (p. 112), but the evidence linking them to MTBE is weak: an editorial, conference papers, a newspaper report and local surveys.
• Thin sourcing and an unweighed trade-off. Load-bearing points rest on drafts and personal communications. The only cost–benefit analysis is Californian. The risk–risk trade-off with benzene and aromatics is raised (pp. 111, 115, 118) but never weighed. There are minor citation slips, e.g. the 1993 existing-substances rules are cited to a new-substances directive (pp. 116, 122–123).
• Positionality. Apparent reviewers overlap with the chapter’s sources (p. 6), and no industry voice is visible.
• Balance. The chapter does concede that MTBE beat lead, that no obvious harm was known at adoption, that EU exposure was lower and that costs might fall.