1) Introduction
In Bellshill Property Limited v (First) Borland Insurance Limited; (Second) Tradebe Healthcare National Limited; and (Third) Grupo Tradebe Medioambiente SL
[2026] CSOH 71 (Outer House, Court of Session), Lord Sandison considered—at debate—whether the pursuer’s case against the
tenant (and the guarantor of the tenant’s lease obligations) was relevant and should proceed to proof.
The pursuer owned a large industrial unit destroyed by fire in July 2021. The first defender (an insurance broker) arranged insurance.
The second defender (tenant) processed clinical waste and, without informing the pursuer, changed its process in March 2021 to use gas-fired
burners to dry waste; one such dryer allegedly caused the fire. The third defender guaranteed the tenant’s lease obligations.
The central issues at debate were: (i) whether the tenant’s alleged lease breaches could have caused the pursuer’s loss (the absence of an
insurance payout), given that fire cover had already lapsed and the insurer later avoided the policy; (ii) how the policy’s “non-invalidation”
and related clauses operated where risk increased without the insured’s knowledge; and (iii) whether the tenant could still be potentially liable
on a counterfactual case that notice would have triggered alternative valid cover.
This opinion superseded [2026] CSOH 70.
3) Analysis
A. Precedents Cited and Their Role
(i) Lease insurance schemes and tenant immunity themes
The second and third defenders relied on Mark Rowlands Ltd v Berni Inns Ltd [1986] 1 QB 211 to frame the lease as an
“insurance scheme” where the landlord looks to insurance proceeds rather than tenant damages for fire loss. Lord Sandison accepted that the
parties’ insurance obligations formed a contractual scheme (para [50]), but the debate ultimately turned less on tenant “immunity” and more on
the causal question: whether the tenant’s conduct could have made any difference given that fire cover had already lapsed and the policy
was avoidable for the pursuer’s non-disclosure.
(ii) Construction of insurance contracts
The pursuer cited Re United London and Scottish Insurance Co Ltd [1915] 2 Ch 167 and Financial Conduct Authority v Arch Insurance (UK) Limited [2021] UKSC 1
for modern purposive/contextual construction. Lord Sandison endorsed a strongly contextual approach, but also criticised the policy wording as generic,
poorly drafted, and ill-suited to fine-grained textual parsing—making context and purpose decisive. (Para [42])
The pursuer also cited Glenfiddich Wind Ltd v Dorenell Windfarm Ltd [2025] CSOH 62 for objective interpretation with background knowledge.
That approach underpinned the judge’s reading of Clauses 3 and 4 as risk-allocation devices protecting an insured owner against unknown occupier acts.
(iii) Non-invalidation clauses and “unknown increase in risk”
The second and third defenders relied on The Seashell of Lisson Grove Ltd v Aviva Insurance Ltd [2011] EWHC 1761 (Comm) and
Ansari v New India Assurance Limited [2009] EWCA Civ 93 to support the proposition that non-invalidation wording prevents an insured’s
interest being prejudiced by unknown occupier acts. Lord Sandison adopted the key thrust of Seashell of Lisson Grove—including rejecting
any requirement that the insured must have discovered/notified the insurer before the loss. (Paras [44]–[46])
The pursuer attempted to distinguish or narrow the non-invalidation effect (arguing, in substance, that it only operates if notice and any additional
premium happen pre-loss). Lord Sandison disagreed, holding there was no textual warrant and that such a reading would drain the clause of commercial utility,
because the “unknown risk” must lie with someone during the period it is unknown—commercially, that is what property insurance is for, provided fair presentation
is made at inception/last review and disclosure is not withheld once knowledge is acquired. (Para [45])
Textbook authority (Colinvaux's Law of Insurance) and older case law (including Greenock Steamship Co v Maritime Insurance Co Ltd [1903] 1 KB 367)
were deployed in submissions about the nature and limits of such clauses, but the decision turns on Lord Sandison’s purposive reading: the clauses allocate the risk
of unknown occupier conduct, while preserving insurer rights to cancel prospectively once notified (here, general condition 15). (Para [46])
(iv) Causation with multiple wrongs: discussion but limited decisional role
The pursuer invoked multiple-cause authorities and commentary—Financial Conduct Authority v Arch Insurance (UK) Limited (including the “two fires” example),
Heskell v Continental Express Ltd [1950] 1 All ER 1033, Baker v Willoughby [1970] AC 467, Williams v The Bermuda Hospitals Board [2016] UKPC 4,
and the line via Wardlaw v Bonnington Casting Ltd 1956 SC (HL) 26 and McGhee v National Coal Board [1973] 1 WLR 1, contrasted with
Performance Cars Ltd v Abraham [1962] 1 QB 33 and Hotson v East Berkshire Health Authority [1987] AC 750.
Lord Sandison did not decide the case on a general relaxation of “but for” causation between broker and tenant. Instead, he undertook a
sequence-of-events analysis to identify the legal position at March 2021 and conclude that—on the actual policy—
the tenant’s conduct could not have caused the absence of cover because fire cover had already ceased and ab initio avoidance was available for composite panels. (Paras [36]–[47])
The causation discussion remained relevant only to the residual counterfactual claim: the tenant’s breach might have triggered steps leading to alternative cover.
(v) Novus actus and scope-of-duty authorities (raised but not determinative)
The second and third defenders cited a range of authorities (including Home Office v Dorset Yacht Co Ltd [1970] AC 1004, Maloco v Littlewoods Organisation Ltd 1987 SC (HL) 37,
and others) in submissions about whether a third party’s act can break causation (novus actus interveniens) and whether the broker’s duty is to protect against such intervention.
Lord Sandison’s decision did not turn on these, because he concluded that—under the actual policy—the tenant’s act did not legally affect cover in the first place, given the prior lapse and ab initio avoidance. (Para [47])
B. The Court’s Legal Reasoning
(i) Identify the loss and fix the timeline
Lord Sandison treated it as “appropriate, indeed necessary” to analyse the precise sequence of events and their legal effects to test causal relevance. (Para [36])
The key pivot was March 2021 (when the tenant changed processes). The judge asked: what insurance position existed then?
(ii) Fire cover had already lapsed: condition precedent endorsement (PD 85)
The electrical inspection endorsement made inspection within two months a condition precedent to liability for fire/lightning/explosion.
Non-compliance by 7 February 2021 meant that, by March 2021, there was no fire cover to be invalidated by the tenant’s later conduct. (Para [36])
(iii) Ab initio avoidance for composite panels (PD 7) was available and unrelated to the tenant
Separately, the insurer was “certainly entitled” to avoid the policy ab initio for non-disclosure of composite panels, given PD 7. (Para [37])
This further cemented that the tenant’s March 2021 process change could not, on the actual policy, be said to have caused the absence of cover.
(iv) Gas-burner non-disclosure was not a valid ab initio avoidance ground
The insurer’s letter also cited non-disclosure of gas burners as an avoidance ground, but Lord Sandison held this was not valid for ab initio avoidance:
burners were not used at inception and there was no renewal/variation; other clauses (Change of Risk, Alteration) would not support ab initio avoidance in these circumstances,
and must be read with the policy’s non-invalidation architecture. (Para [39])
(v) Exclusion 8 did not remove cover for a building fire merely because heat was used in the tenant’s operations
Lord Sandison rejected the argument that Exclusion 8 barred the claim because the fire arose from a process “involving the application of heat”.
The exclusion concerned damage to the part of the insured property that was itself being heated/processed. Here, waste—not the building—was being heated. (Para [40])
(vi) Clauses 3 and 4: purposive risk allocation; no “pre-loss notice” requirement
The judge interpreted Clauses 3 and 4 as allocating the risk of unknown occupier acts/omissions to the insurer, protecting the insured’s interest
against policy invalidation while the matter remains unknown. (Paras [41]–[45])
Critically, he held Clause 4 does not require the insured to discover and notify the issue before damage occurs; otherwise the clause
would lack meaningful commercial purpose. Once the issue becomes known, the insurer’s protection lies in prospective cancellation (general condition 15),
not retrospective avoidance. (Paras [45]–[46])
(vii) Therefore: the tenant did not cause the lack of cover under the actual policy
Combining the above, Lord Sandison held the absence of cover for the fire loss under the policy actually in place was not caused or contributed to by the tenant:
fire cover had already ceased; and ab initio avoidance was available for composite panels irrespective of tenant conduct. (Para [47])
(viii) Residual relevance: a counterfactual “alternative insurance” claim via lease clause 1.3
The pursuer could only proceed against the tenant if it could prove that tenant compliance would have set in motion steps leading to new or amended valid cover by the time of the fire. (Paras [48]–[49])
Lord Sandison analysed the lease duties relied on: clause 1.2 (do nothing to invalidate insurance/comply with insurer requirements) and clause 1.3 (do not use premises in a way that may make additional premium payable without agreeing to pay it).
He considered that clause 1.2 was not sensibly engaged on these facts (no insurer “requirements” about burners; and “invalidating” the insurance in place was not established). (Para [51])
Clause 1.3, however, was arguably breached: using gas burners could be viewed as a use that “may make” an additional premium payable, triggering a duty to agree to pay it. (Para [52])
Causation of loss would be difficult, requiring proof of a chain of counterfactual steps (notice; enquiry; market willingness; new policy terms; compliance with precautions; premium payment; and that a fire still occurs and is covered). (Para [53])
But difficulty of proof was not a reason to strike the case out at debate; the pleadings were just sufficient for proof before answer. (Paras [53]–[54])
(ix) Mutuality: not a simple “landlord breach extinguishes tenant duties” rule
Although the landlord (pursuer) had breached its own insurance-maintenance obligation by allowing fire cover to lapse, Lord Sandison rejected an over-broad mutuality argument that the tenant’s insurance-related duties were wholly suspended.
The better approach was a contextual construction: tenant duties remain meaningful in relation to whatever insurance cover is actually in place from time to time, even if one head of cover (fire) has fallen away. (Para [50])
C. Impact and Significance
(i) Non-invalidation clauses: protection can apply even where loss precedes discovery
The clearest practical point is the court’s endorsement of The Seashell of Lisson Grove Ltd v Aviva Insurance Ltd on timing: non-invalidation protection is not confined to cases where the insured notifies the insurer before the loss.
That matters for landlords/owners who are not in day-to-day control of premises and for coverage disputes arising from undisclosed tenant activities.
(ii) Insurance causation disputes in mixed-defender litigation
The decision illustrates a structured method for multi-party disputes: identify the insured’s loss (here, absence of an insurance payout), then determine—by reference to endorsements and avoidance—whether a co-defendant’s alleged breach could
as a matter of law have altered the coverage position. Where cover has already lapsed, later tenant conduct may be causally irrelevant to the coverage deficit, even if it caused the fire.
(iii) Lease drafting and operational change governance
The survival of the clause 1.3 argument signals that “additional premium” clauses may generate meaningful obligations to disclose operational changes—supporting landlord claims framed as “lost opportunity to procure alternative cover,”
even where the existing policy is defective for reasons attributable to the landlord/broker.
(iv) Litigation management: proof before answer where counterfactual insurance markets are in issue
Lord Sandison accepted that the counterfactual chain would be hard to prove, but allowed it to proceed. Parties in similar cases should expect the court to require evidence about market availability, underwriting appetite, risk-control measures,
and premium adjustments rather than resolving such issues purely at the pleadings stage.