RSA Fresh Water Flood Coverage Acknowledges "Concurrent Causation" in Flood Damage Claims

An article in candianunderwriters.ca describes the new RSA fresh water endorsement expected to be available to the large majority of RSA customers.  It is encouraging to see the recognition that there has been 'concurrent causation' for flood damages:

"Some of the challenge we had with the floods in Alberta and in Toronto was this concurrent causation issue, where you had sewer backup happening at the same time as you had clean water flood coming in and you can't really say 'Part of it is brown water, part of it is clear water,'" the RSA representative said.

This represents progress in terms of characterizing the complexity of flood risks, but what is missing from the discussion (for Toronto table land type flooding as opposed to Alberta riverine flooding) is that:

Extraneous clear flow from groundwater infiltration
and inflow points enters the 'brown' sanitary sewer.

Comment 1) Sewer backup water during severe storms is often brown and clear together, because during extreme storms, the extraneous inflows to the sanitary sewer system dilute the typical brown water, typically by a factor of 10 to 1.  For infrastructure managers, the clear water components is often called RDII, or rainfall dependent inflow and infiltration - this clean water component can peak at up to 5 L/s/ha in a partially separated sewer system during an extreme event. Because of this, separating clear water and brown water risks is always difficult to do in practice.

Clear water surface flooding enters doors and windows and
exits low lying  upstream properties via floor drain.

Comment 2) The clear and brown water coming in typically occurs at different locations in the neighbourhood's drainage system.  Low-lying areas with low exposed openings can allow clear water to enter buildings and then exits via floor drains.  Those extraneous flows enter the sanitary sewer system and cause brown+clear water to enter downstream properties via floor drains.  Because of this, a downstream property may be vulnerable to brown water backup, because of an upstream property's clear water risk.

RSA indicated that regarding the old sewer backup coverage:

Downstream sanitary sewer overwhelmed with brown and
clear water creates backup, entering property via floor drain.

"In the past, that endorsement was much broader and in some respects would cover flood not intentionally but did cover flood."

Comment 3) What this means is that for some insurers, the backup premiums may have been increasing to pay for clear water, overland flood damages.  In my case, for example with a different insurer, the premium has increased 800% and the coverage limit has decreased for backup insurance - this could be explained by unintentional coverage of clear water, overland flooding in my city / neighbourhood. In an ideal world, RSA could decrease backup premiums by limiting coverage of uninsured perils like clear water, overland flooding, while pricing clear water flooding separately. .
Because of Comment 1 and 2, it will difficult to separate the different risk coverage

Just like cable TV, you can now pick and pay for your flood peril coverage:

"If you get Waterproof you get sewer backup and the flood, but if you don't want the Waterproof endorsement, if you don't want the flood, you can just buy limited sewer backup," RSA adds, saying that they are "putting a lot of emphasis" on training brokers.

Comment 4) This approach may meet RSA goals of limiting coverage for uninsured perils and increasing premiums for flood to help cover increasing costs during extreme events.  In the end, the bottom line for the business is not dependent on having an optimal, precise vulnerability assessment for individual properties - all it needs is a sub-optimal net positive outcome on the portfolio.  RSA is making an incremental adjustment toward a more sustainable, better priced risk model with the changes.

Not all properties are eligible for flood coverage though:

RSA noted  that for high risk properties, it would be difficult to cover fresh water flooding "because there is almost a guarantee that something is going to happen," she added.

Comment 5) High risk properties can include those in a defined river floodplains.  It is reasonable approach that coverage is not available and instead land use policies should deter such high risk developments over time.  What is missing in this approach is that governments and their agencies (Conservation Authorities in Ontario) and insurance companies are not managing all risks in a complete and comprehensive way.  Will RSA risk zones be shared with municipalities, or with property owners to support land use planning policies or flood proofing efforts by individuals or businesses? Not likely. Or will the sole purpose of the risk maps be to make decisions on the portfolio for the benefit of shareholders - that is likely - there is nothing wrong with that, but if so, let's take a step back and consider what the insurance industry's role really is in terms of managing society's flood risks.  Are the efforts by RSA to define flood risk a benefit to society or the balance sheet? Does it make sense that individual insurers redo overland flood risk assessments independently from each other, or should a public flood underwriter make those assessments to be shared by the insurance industry and municipalities?  In that manner, there would be no gaps in coverage and risk mitigation strategy.

Comment 6) What is needed to enhance the RSA sub-optimal, incremental improvement to risk pricing for flooding during extreme storms? The answer is a more robust, neighbourhood scale vulnerability model and a comprehensive approach to riverine flood risk.  This could result in and mandatory coverage for highest risk overland flood properties as one outcome.  Similarly, premiums for backup coverage could be increased in neighbourhoods with high overland flood risk, because clear water entering a few properties and overwhelm the sanitary sewer system in a widespread area. Our research shows how overland flood and backup (basement) flood risks are correlated in this post. Because risks are at a neighbourhood scale and policies are written at a property scale there is no way that the current insurance model can be optimal.

RSA notes in the article that "Canada doesn't have a lot of high risk zones with respect to flood."

Comment 7) In making this statement RSA should have distinguished between urban overland flash flooding and riverine flooding in a valley systems.  Certainly a low percentage of properties are within vulnerable valley flood plain areas where rivers and creeks can swell.  But on table land, urban flash flooding affects many neighbourhoods built before improved 1980's drainage standard improvements. These table land areas are where 'concurrent causation' really occurs.

Lastly RSA notes "That is going to be very clearly defined as what's covered and not covered"

Comment 8) Costs and revenue may be in a more sustainable balance with the RSA policy updates, but this may not necessarily be because policy coverage is more 'clear' in the future. As noted in the initial comments, there will generally always be some 'brown' mixed into the system when it comes to extreme storms.

Overland Flood Factors Affect Basement Flood Incidents - Correlating Insurance Perils and Mapping Risk

Overland flow risk mapping and basement
flood history highlights risks on table land,
outside river flood plains and traditional hazard
mapping areas.

Flood is the new fire.

Blue is the new red ink, as water damage claims increase. Enough slogans though ... time for insights resulting from new overland risk mapping for the GTA!

Overland flooding had recently becoming an insured peril for residential property owners, offered by Aviva Canada.  Others are following suit.  So now is the perfect time to explore the differences in "surface flooding" types in an urban environment, to assess risk factors and even look for correlations with traditional insured perils, like sewer back-ups that cause basement flooding.  Understanding risk factors can support better decisions and help direct mitigation efforts.

First some set-up and definitions.

Media often reports the devastating flooding in Alberta and Toronto in 2013 in the same breath, but there are important distinctions to be made - High River's "riverine flooding" along the low-lying bottom of a defined valley or wide flood plain is quite different than "urban flooding" high up on the table land (i.e., many thousands of Toronto's flooded homes).  To put things in perspective, the Toronto's Don River has a drainage area of 360 sq.km contributing to valley flow, while a table land area subject to urban flooding may have as little as a 0.01 sq.km, or 10 hectare drainage area.

Riverine flooding is largely manageable - spatially fixed, quantifiable in terms of flow frequency and flood depth, and (in Ontario) regulated through the provincial policy statement on natural hazards, MNR guidelines, and conservation authority regulations.  In the past, flood plain mapping to define riverine flood risks was once the 'bread and butter' of engineering consultants, who started number-crunching U.S. Army Corps of Engineer's HEC-2 simulation models back using punch cards - yes, that long ago.  The limitation is that only a few percent of Ontarians live in flood plains according to Conservation Ontario.  And based on new analysis from cityfloodmap.com, only about 2% of May 12, 2000, August 19, 2005, and July 8, 2013 Toronto flooding occurred in river flood vulnerable areas!

Urban flooding, in contrast, affects a greater proportion of the population - 86% of Ontarians live in urban areas per the 2011 census - but this type of flooding is not as readily managed.  In fact, risk factors equivalent to flood plain maps are seldom ever mapped.  This is because urban flooding is transient and overland flow paths are not permanent physical features that are part of i) a defined valley, or ii) municipal drainage infrastructure.  Overland flow paths appear for a few minutes or hours every few decades and then subside.  So calling this 'flash flooding' is appropriate - engineering studies show severe urban flooding events are most highly correlated with high short-term rainfall intensities over minutes.  Hours later, the overland flow is gone.

To prevent urban flooding, engineers have been designing the overland drainage network though subdivision grading since the early 1980's when "dual drainage design" emerged in Ontario.  This approach recognized the limited capacity of the minor system (underground sewer system) for conveniently conveying small storm runoff, and the need to design the major system (above-ground overland system) for extreme storms.  In pre-1980 development, the major system still exists but it can run haphazardly beyond roads, through back yards, between houses and into basement windows and walkouts. And this critical, non-designed, transient overland flood risk has not ever been mapped - ta da! - until now.

New analysis from cityfloodmap.com in the slides below ties overland flood risks all together, reviewing flood incidents across regulated valley hazard areas and river flood vulnerable areas (riverine flooding), and within overland flow risk areas across Toronto table lands (urban flooding).  It provides insight into risk factors affecting overland flooding and direction on de-risk opportunities.  Most importantly, it demonstrates the correlation between traditional sewer back-up peril and the emerging overland peril, which should cause some insurance providers to evaluate portfolio risk (where should coverage be coupled or excluded in isolation, recognizing the correlations).

Urban Flood Risk from Flood Plains to Floor Drains from Robert Muir

If flood is the new fire for the insurance industry, this new analysis, which demonstrates neighbourhood scale flood risk factors due to overland and catchment slope risks, should also prompt discussion on how flood risks are assessed, priced and prioritized for mitigation, adaptation and insurance purposes. Just as fire underwriters identify neighbourhood scale risk factors and consider fire suppression capabilities of municipalities, flood risk assessors should consider the overland risk factors, that extend well beyond the extent of individual properties. Determining risks should not only consider whether an individual property has a reverse slope driveway when broader overland flood risk factors can be readily quantified.