Global Environmental Regulation, Policy, M&A & Market Update

WSP presses a twice-rejected pursuit of Arcadis.

• The chronology: WSP Global submitted an initial non-binding proposal to acquire all of Arcadis N.V. for €48.50 per share on July 1 — rejected unanimously by Arcadis' Executive and Supervisory Boards on July 14 as not reflecting the company's intrinsic value, strategic position, or future prospects. WSP came back on July 23 with a revised proposal at €51.50 per share (~€4.5 billion total), which Arcadis' boards again unanimously rejected on July 30.^[1],[3]
• Deal structure: the revised offer represents a 45.8% premium over Arcadis' unaffected July 22 closing price of €35.32, structured as roughly 50% cash and 50% WSP stock, with shareholders able to elect either option subject to proration. WSP explicitly invited Arcadis' largest shareholder — the employee foundation Stichting Lovinklaan, which holds a 19% stake — plus Katalys, to serve as reference shareholders in a friendly transaction.^[1],[4]
• Where it stands: WSP says no definitive agreement has been reached and any transaction remains subject to the two companies agreeing on terms; Arcadis has flagged concerns beyond price, including strategic fit, cultural fit, and deal certainty. One sell-side analyst called blending WSP's serial-acquisition culture with Lovinklaan's continuity mission for Arcadis "not easy."^[3],[4]
• The backdrop: WSP reported record Q2 2026 results on Aug. 5 — revenue of $5.40 billion (up 19.9%) and a record $20.1 billion backlog (up 23.2%) — with growth substantially driven by its own recent acquisitions of TRC (February 2026) and Ricardo (October 2025). On the earnings call, management acknowledged the Arcadis pursuit "introduces uncertainty" but reiterated its intent to keep pursuing a friendly, recommended deal.^[4]

Clean Harbors expands field services and its data-center offering.

• The ES&H acquisition: announced with Q2 earnings on July 29, Clean Harbors signed a definitive agreement to acquire Louisiana-based ES&H for $305 million in cash. ES&H operates 13 service branches across Louisiana and Texas, holds the Coast Guard's highest Oil Spill Response Organization classification, and is expected to contribute ~$90 million in annual revenue and ~$30 million in adjusted EBITDA — implying roughly an 8.7x post-synergy multiple after $5 million in expected cost synergies. Closing is targeted for the second half of 2026.^[5],[7]
• The manufacturing contract: Clean Harbors also closed a 10-year, ~$600 million disposal agreement with an unnamed manufacturing customer expanding U.S. operations through 2030 — management compared it to a prior contract with 3M, tying it to broader reshoring trends. The deal begins in Q4 2026 and is expected to reach an $80–100 million annual run rate by 2030.^[6]
• The data-center build-out: separately, Clean Harbors is spending roughly $50 million on specialty equipment, tankage, and vehicles to serve data-center customers, pitching mechanical flushing, chemical passivation, and water filtration services for facilities in active construction phases.^[5]

Arcwood Environmental closes its sixth roll-up in two years.

• The deal: announced Aug. 5, EQT Infrastructure-backed Arcwood (formerly Heritage Environmental Services) acquired Rockville, Maryland-based Environmental Management Services, Inc. (EMSI) — a 42-person hazardous and regulated medical waste manager serving healthcare systems, pharmaceutical manufacturers, and government agencies.^[8],[9]
• The pattern: this is Arcwood's sixth acquisition since its 2024 rebrand — following January's MXI Environmental Services/Dynamic Recycling deal (Virginia) and May's Safeway acquisition (Alabama) — part of "hundreds of millions of dollars" in cumulative deal spend plus an $80 million 2025 facility-upgrade program (new incineration capacity in Texas, shredding in Ohio).^[8]

RFS Small Refinery Exemptions: EPA finalizes six more decisions.

• The action: on August 3, EPA decided six pending small refinery exemption (SRE) petitions from four refineries for the 2023–2024 compliance years — granting one full (100%) exemption, two partial (50%) exemptions, denying zero, and finding three petitions ineligible under Clean Air Act §211(o)(9).^[10],[11]
• Scale: the decision exempts roughly 160 million Renewable Identification Numbers (RINs) for the 2024 compliance year. Thirty-six SRE petitions remain pending.^[11]
• Reaction: Growth Energy, the biofuel industry's largest trade group, said it is "closely evaluating" the decisions to ensure they don't undermine biofuel, agricultural, and rural-economy investment made on the strength of the RFS.^[11]

California SB 253: the Aug. 10 deadline was never really the deadline.

• What's happening: on July 27, CARB released proposed modifications deferring the first-year Scope 1 and Scope 2 GHG emissions reporting deadline from August 10 to November 10, 2026, alongside first-year enforcement relief. The proposal is out for a 15-day public comment period closing August 11.^[12]
• The catch worth flagging: because the deferral is itself only proposed, reporting entities technically face regulatory uncertainty right through the original deadline — CARB's enforcement posture already emphasizes discretion for "good-faith first-year submissions," which should reduce near-term compliance risk regardless of the modification's final status.^[12],[13]
• Fees: CARB anticipates assessing flat fees — estimated at $3,106 per entity for SB 253 and $1,403 per entity for SB 261 — on September 10, 2026.^[13]

Background context: two deregulatory proposals continue moving through the pipeline. EPA's "Common Sense Approach to Chemical Accident Prevention" rule — which would roll back 2024 Safer Technologies and Alternatives Analysis (STAA) requirements for existing petroleum and chemical facilities, retaining them only for newly built processes — was published February 24 with comments closing May 11; EPA's own estimate puts annualized savings at $234.7–240.3 million. No finalization date has been announced.^[17],[18] Separately, EPA's July 9 proposal to eliminate forced speed deratements and automatic shutdowns triggered by DEF sensor failures on MY2027+ heavy-duty engines — which we covered in our July 10 issue — remains in its comment period, with EPA projecting roughly $12 billion in fleet operator savings.^[19]

• Hyperscaler load growth: Clean Harbors' new data-center offering (Section 01) is a direct read on this trend — rapid AI and data-center expansion is driving record demand for specialized industrial environmental services spanning cooling-water sourcing, thermal discharge compliance, and dedicated hazardous byproduct handling, with incumbents building out dedicated equipment and capacity rather than treating it as incremental volume on existing infrastructure.
• Zero Liquid Discharge (ZLD) adoption: stricter discharge limits and rising municipal intake tariffs are converting industrial wastewater management from a pure compliance cost into a strategic asset — the same dynamic underpinning DuPont's FilmTec XC220 recognition this week (Section 04), where higher brine concentration directly reduces downstream treatment volumes and cost.
• Extended Producer Responsibility (EPR) mandates: implementation continues expanding across states with active EPR laws, requiring packaging, textile, and electronics companies to finance and structure formal end-of-life collection and recycling channels — a durable, non-discretionary compliance category that keeps generating recurring consulting and program-management work independent of any single state's political cycle.

DuPont's FilmTec XC220 gets recognized — four months after launch.

• This week's news: on August 5, DuPont announced its FilmTec Fortilife XC220 element was named a Sustainability Product of the Year by the Business Intelligence Group.^[14]
• What the product does: launched in April 2026 as part of DuPont's broader Fortilife portfolio expansion, the XC220 is a high-pressure reverse osmosis element that achieves brine concentrations up to 220 g/L NaCl at operating pressures up to 83 bar — extending membrane-based treatment into concentration ranges traditionally requiring energy-intensive thermal evaporators for Minimal and Zero Liquid Discharge systems.^[15]
• Why it matters commercially: replacing thermal evaporation with membrane-based concentration lowers both capital and operating costs for industrial ZLD/MLD systems — directly relevant to the ZLD adoption trend in Section 03.

AI-optimized wastewater treatment: a genuinely rigorous number just landed.

• The new study: a quasi-experimental analysis of AI aeration-control deployment across French wastewater treatment plants — exploiting variation in adoption timing and outages to isolate causal effects — found full-time AI control reduces plant electricity consumption by 5.4% and carbon emissions by 6%, with energy expenditures falling 8.2% (the added electricity draw from running the AI models itself is under 1% of the savings). Effluent water quality improved alongside the energy savings.^[16]
• How this compares to vendor claims: industry and vendor-published figures for AI aeration control have circulated in the 18–30% energy-reduction range; this causal, quasi-experimental study is more conservative but also far more methodologically rigorous — a useful benchmark for underwriting realistic savings rather than best-case vendor pitches.^[16]

Two large, overlapping platforms — and a real strategic logic for combining them.

• WSP and Arcadis are both top-tier global "data-driven sustainable design, engineering, and consultancy" firms with heavily overlapping environmental, water, transportation, and buildings practices — precisely the categories this newsletter tracks weekly for permitting-scarcity and compliance-driven demand.
• WSP's own investor materials describe the potential combination as accretive on multiple dimensions: high-single-digit percentage accretive to adjusted EPS before synergies, mid-teens percentage accretive once synergies are realized, and additionally "geographic, data, domain expertise, client and capability accretive" — WSP's language for the deal expanding its footprint and technical bench rather than simply adding revenue.

Why Arcadis keeps saying no.

• Arcadis' boards have now unanimously rejected two unsolicited proposals — €48.50/share on July 14, and the revised €51.50/share offer on July 30 — citing not just valuation (the boards say the price doesn't reflect Arcadis' "intrinsic value, strategic position and future prospects") but also strategic fit, cultural fit, deal certainty, and the interests of other stakeholders.^[2]
• The employee-foundation ownership structure is a real obstacle, not just an analyst talking point: Stichting Lovinklaan's 19% stake and its "continuity mission" sit awkwardly against WSP's history of serial, acquisition-led growth — a cultural mismatch that a higher price alone may not resolve.^[4]

WSP isn't backing off. The Aug. 5 Q2 results — record $20.1 billion backlog, 19.9% revenue growth, and management publicly reiterating intent to pursue a friendly transaction despite two rejections — suggest WSP has both the balance sheet and the stated appetite to keep pressing, potentially with a further-improved offer or a longer negotiation aimed at winning over Lovinklaan specifically.

What the offer implies on an EV/EBITDA basis — and how it stacks up against WSP's own recent deals. Neither WSP nor Arcadis has disclosed an EV/EBITDA multiple for this proposal (unlike WSP's Ricardo and TRC deals below, where WSP published the multiple itself), so we've built one from disclosed figures:

• Equity value: ~€4.5 billion, per WSP's own disclosed offer (€51.50/share).^[1]
• Plus net debt: €1,042 million, per Arcadis' H1 2026 results reported the same day as the second rejection (July 30) — giving an implied enterprise value of roughly €5.5 billion.^[21]
• Divided by estimated EBITDA: Arcadis discloses a Net Debt/Operating EBITDA leverage ratio of 1.5x on FY2025 net debt of €797 million, implying FY2025 Operating EBITDA of approximately €530 million — consistent with the 13.8%–14.0% Operating EBITDA margins Arcadis has reported across recent quarters on ~€3.8 billion of annual net revenue.^[21],[22]
• Implied multiple: approximately 10.3x–10.5x trailing EBITDA — a Gaya Capital estimate, not a company-disclosed figure, and sensitive to which EBITDA period and net debt snapshot is used.

For broader context, general engineering-consulting M&A benchmarks put "premium scale" platforms — multi-state, federal/infrastructure-exposed, $25M+ EBITDA — in the 10x–13x+ range in 2026, with mid-size multi-state platforms at 8x–11x and smaller regional practices considerably lower.^[23] Our implied WSP/Arcadis multiple sits at the lower end of that premium-scale band, and almost exactly in line with WSP's own Ricardo acquisition — but meaningfully below the 14.5x/12.5x WSP paid for TRC, its highly sought-after U.S. Power & Energy platform, just months earlier.

This week's clearest signal is that scale and contracted demand are pulling environmental-services investment in the same direction: toward the largest, most diversified, permit-rich platforms in each category. WSP's aggressive, twice-repeated pursuit of Arcadis, Clean Harbors' targeted data-center and manufacturing capital deployment, and Arcwood's steady regional roll-up all fit that pattern, even though they look like three unrelated stories on the surface.

On the regulatory side, this week is a reminder to read compliance deadlines carefully before building timelines around them — California's SB 253 "deadline" was already softened by CARB's enforcement posture well before the formal deferral proposal, and the practical compliance-software and assurance demand underneath the headline date doesn't change regardless of which date ultimately sticks. The RFS exemption backlog remains a slow-moving but persistent source of RIN-market uncertainty, with 36 petitions still pending after this week's action.

In technology, this week offered a useful lesson in evidence quality: a rigorous causal study putting AI aeration savings at a conservative 5.4–8.2% is more informative for underwriting than higher vendor-claimed figures, even though the vendor numbers make for better marketing. As AI-enabled water and wastewater technology matures from pilot to mainstream deployment, expect more of this kind of rigorous, population-level evidence to emerge — and expect it to matter more to sophisticated buyers than best-case pilot results.