The Marsh Institute has officially published a report in 2020 as” 100 Largest Losses in the Hydrocarbon Industry, 1974-2019". This report has made very interesting connections about insurance risk management with process safety.
Ref: https://www.marsh.com/us/industries/energy-and-power/insights/100-largest-losses-in-the-hydrocarbon-industry.html
In the introduction of this report, regarding past mistakes, the following is well presented:

Past Mistakes
The most commonly cited energy industry risk recommendations made by Marsh JLT Specialty over the past few years (“systems of work,” “inspection,” and “fireproofing”)
were prevalent issues among the largest issues in 1988-99 — reinforcing the view that the industry may not have learnt from past mistakes.
During the Piper Alpha incident in 1988, the largest industry property damage loss of all time, there were shortcomings in all three areas of the root cause “systems of work” (which includes permitting, management of change, and shift handover).
Similarly, the largest losses in 1988-1989 and 2018-19 both suffered from the escalating factors “emergency response plans” (ERPs) and “fire protection.” For example, the lack of adequate ERPs caused the initial incident to escalate at Piper Alpha; at Campos Basin it took over a month to control the resulting fire, compounding the cost of the incident.
In 2018-19, in all four of the largest losses, the sites were not well prepared for the incidents that occurred. And in 2019 the US$600 million explosion at Limbe, Cameroon,
began as a fire near a distillation unit on the refinery, which escalated to a much larger explosion. More robust fixed-fire protection at site may have reduced or prevented this
escalation.
Four key factors arguably prevent lessons from being learnt from losses:
-
Distance: Parties unconsciously feel less affected by events a long way away.
-
Culture: Preventing lessons from being implemented effectively.
-
Tunnel vision: Not realizing wider relevance of lessons.
-
Time: Lessons are learnt, but then forgotten or solutions are insufficiently robust.
Given the 30 years that have elapsed between the two-worst periods in terms of the 20-largest losses, time would seem to be a particularly significant factor.
At AIFCO, we have integrated planning to use this report in our engineering and consultant projects. Services that, while improving process safety and production stability, will greatly help to reduce the costs of insurance for your facilities and critical equipment.