Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
8 TREATMENT OF LEVEES IN THE INSURANCE ASPECTS OF THE NFIP DISCUSSION AND RECOMMENDED APPROACH Currently, the crediting of levees for setting flood insurance premiums is based strictly on the level of protection provided, with the 100-year flood level being the dividing line between "all or nothing" credit. For properties behind a levee equal to or higher than the crest of the 100-year flood, no premium is added (accounting for hydrologic uncertainties or the possibility of structural failure during lesser events) to the basic rate charged insurees outside a marked floodplain. Conversely, insurees behind levees failing to meet the benchmark level of protection are given no reduction in their insurance premiums that recognizes the levee's presence. The committee recommends: ⢠Regardless of the level of protection provided, the levee-protected area should be disaggregated into flood risk zones and an actuarial rate be established for each zone that reflects the degree of protection actually provided by the levees. This ideal would entail the development of a practical procedure for estimating the probability of failure on an annual basis, the actuarial chance of loss, for properties protected by specific groupings of levees. Premiums based on actuarial risk would require individuals who own existing or who construct new buildings in areas protected by levees to make equitable contributions to the cost of the loss-sharing mechanism arranged on their behalf by FEMA. Each policyholder for property behind a levee would pay a premium determined by a reasonable estimate of the probable residual flood damage, recognizing the inherent uncertainty in the estimate. Precedent for a rating scheme based on approximate average group risk is found in the fire insurance industry in which assignment of property to 1 of 10 fire risk rating zones is according to a determined degree of exposure. Windstorm insurance rates are regionalized, sometimes by state or by proximity to the South Atlantic or Gulf Coast. Properties in "tornado alley" are called upon to pay higher premiums than those elsewhere. -39-
It is in the best interest of the public and FEMA to encourage communities and individuals to take actions that mitigate flood damage to the extent practical. One workable incentive is to reduce the insurance rates paid by residents whose communities provide levee protection. The reduction provides residents a tangible incentive to pressure their communities to provide levee protection. Later, should a levee owner fail to maintain a levee to the degree required for continued FEMA recognition, the residents would be notified that their flood insurance rates would be set to higher levels. Many would respond by applying pressure to their communities to correct the situation. Such local pressures tend to be much more effective than those applied directly by the federal government. The committee recommends: ⢠Consistent with the design levels of protection recommended in Chapter 3, NFIP policyholders in areas behind existing levees that offer more than 25-year protection or new levees affording 100-year protection or greater should pay lower rates that reflect the reduced risk of property damage. In order to estimate appropriate insurance rates, levees would be classified by a set of factors carefully prepared by FEMA intended to reflect the probability of failure; that is, the likelihood of failure would be estimated based on correlation with such factors as levee geometry, levee maintenance, and adequacy of internal drainage. Preliminary estimates of the rates of failure for various classifications can be based on historical data from the Corps of Engineers, SCS, FEMA, and such state agencies as the California Department of Water Resources. The committee recommends that all property owners pay full premiums for flood insurance with no credit given for reduced risk because of any levee until the affected community makes application to FEMA. Subsequent to the process of levee inspection, evaluation, and acceptance by FEMA, as described in Chapter 4 and based on factors affecting levee effectiveness, credit for risk reduction would be given as guided by the correlation discussed in the previous paragraph by adjusting the insurance rate accordingly. The community, levee owner, or property owners behind the levee would be expected to pay for the inspection and for having the evaluation. Reports on inspection of maintenance, subject to FEMA verification, should be periodically submitted for approval at the community or levee owner's expense. Failure to demonstrate adequate maintenance would automatically increase insurance rates to a higher classification. The committee recommends: ⢠The levee flood risk zones should match the existing flood risk zones established for the regular Flood Insurance Program. -40-
⢠FEMA should continue its efforts to establish an actuarial rate basis for the Flood Insurance Program and, as far as practical, convert its present rate schedule to actuarial rates. The actuarial rates, as is done by the private insurance industry, would be updated annually to reflect experience, claims paid, and the cost of doing business. A solvent NFIP requires that policyholders be assessed at actuarial rates. The desirability of an actuarial-based rate structure has been stated repeatedly and is endorsed here. FEMA indicates that its current policy is to implement actuarial rates over the next few years. The actuarial rate schedule proposed by FEMA would be designed to make the total program financially self-sufficient, thereby obviating the need for federal subsidy. It is recognized that subsidized rates may have to be phased out over a period of timeâfor example 10 yearsâor under condition of change in property title. Nevertheless, with a policy of reducing federal expenditures, the current subsidized costs should be transferred to those under risk. The actuarial rate schedule adopted by FEMA should be based on the level of risk associated with floodplain development and not be designed solely to make the program financially sound. That is, the rates paid by each policyholder should be a reasonable reflection of the potential risk associated with the individual existing or proposed floodplain development. It is realized that: (1) existing construction (built prior to the time when the first flood insurance rate map became effective) is by statute subsidized to provide for affordable rates, and (2) FEMA or FIA is not free to eliminate this subsidy. Congressional consideration of eliminating this subsidy over time would be appropriate. HAZARD RATING PROCEDURE Background The use of actuarial risk for setting insurance rates requires that a procedure be developed which can differentiate the failure hazard associated with individual levees and assign numerical assessments of hazard. This procedure must relate hazard to identifiable conditions of the levee and the maintenance performed on it. Further, the procedure must, to the extent possible, be based on objective evaluation criteria and be publicly defensible. Given the current state of hydrologic and geotechnical risk assessment, the procedure must be simple in concept, and, given the scale of a levee evaluation program, it must also be simple in execution and administratively feasible. The development of risk assessment procedures for levees is limited by the inherent difficulties in probabilistic risk analysis for -41-
earthen structures and their foundations. While methodologies for evaluation exist, data requirements are extensive and the results, with incomplete or approximate information, may be questionable. Presently, risk assessments can best be based on statistical analysis of the historical record, on expert opinion, or some combination of these. Risk estimates must be based on measurements of levee conditions that have been related to the probability of failure, either by empirical evaluation of the historical record or by engineering analysis. The major problem in employing statistical analysis of empirical data is that so few failures have been recorded. The failures that have been recorded come from a wide range of levee types. When data are disaggregated to estimate the effects of levee design, physical characteristics, soil and hydrological conditions, maintenance history, and other relevant factors, the number of failures in any one category is quite small for reliable statistical analysis. On the other hand, the problem of assigning risk ratings to levees is not unlike many other insurance rating problems, e.g., fire safety ratings. Indicators of levee conditions and probable performance can be identified, and these indicators can be ranked in approximate order of importance. After grouping indicators in a limited number of classes, the probability of structural levee failure associated with various combinations of indicator values can be estimated from failure statistics and expert opinions. An overall risk rating is obtained by combining the estimate of structural risk from indicator information with hydrologic forecasts, and the risk ratings are then related to other risk classifications currently used by FEMA. A rating scheme for areas behind levees is proposed below which incorporates consideration of hydrologic and geotechnical risk and uncertainty. It is suggested that hydrologic risk and uncertainty (depth, duration and frequency of flooding) for areas behind levees be handled by the existing NFIP procedures and rate zones. Hydrologic risk and uncertainty must also be incorporated into new procedures for evaluating geotechnical risk and uncertainty. Geotechnical risk and uncertainty can be considered by providing for lower premiums for owners of floodplain property protected by levees that are adequately designed, constructed, and maintained. This premium reduction would not result if the levee were not properly designed, constructed, and maintained and if a properly designed and maintained interior drainage system did not exist. The following rating scheme is presented as an example. Development of the specific procedures for the geotechnical rating scheme would be the responsibility of a future contractor and is beyond the committee's capability. Rating Scheme The probability of levee failure under flood conditions depends on both hydrologic and geotechnical risk and uncertainties. -42-
The conceptual approach to estimating hydrologic risk is theoretically clear and well established in practice even though hydrologists recognize that a variety of uncertainties constrain our ability to do so precisely. The reasonable way to estimate the hydrologic component of risk to lands behind levees is to employ these methods as done in existing NFIP procedures, while recognizing the inherent imprecision in the results. The conceptual approach to estimating geotechnical risk at this time is theoretically incomplete and therefore not generally applied in practice. Nevertheless, it seems clear that the primary piece of hydrologic information determining risk for a given levee is the probable distribution of duration that water levels can be expected to continue at elevations lower than levee crest. The aggregate probability of structural failure could be found by integrating the conditional probability at given differential heads and duration multiplied by the probability distribution of water level and duration. While probabilistic methods have been developed for some geotechnical risks, and are being developed for others, the geotechnical engineering community has not universally embraced the concept that reasonable quantification of the probability of failure is practical. Certainly, the state-of-art is far from a highly sophisticated quantification scheme. In this situation the committee recommends developing a scoring procedure for evaluating geotechnical hazard from simple indicators that correlate to levee performance. These indicators could include observable items such as design and construction features, embankment geometry, obvious signs of distress, adequacy of maintenance programs, and similar easily identifiable properties. For each indicator, scoring criteria could be developed which assign a score of 0, l, or 2 to each possible condition, depending on whether the condition is inadequate, requires remedial work, or is satisfactory. These numbers can then be aggregated to yield a risk score. If a levee is judged to be unsatisfactory in a key category, then the levee would be unacceptable for reduced insurance rates until the condition is corrected to a satisfactory standaxl. From a practical view, the list of indicators used will have to be limited to a reasonably small number and amenable to evaluation. This should facilitate evaluation of individual levees and help organize a statistical data base to form a basis for improved estimation. The suggested beginning for a procedure for incorporating geotechnical risk and uncertainty into the rating procedure (Tables 2 and 3) would result in a score, S_t which would determine whether a reduction would be made to the insurance premium as a result of geotechnical risk and uncertainty. If the estimated value of £> exceeds various values, then various reductions would be subtracted from the premium. The premium from which the reduction is subtracted is that premium established by NFIP procedures for the conditions of hydrologic risk in existence for the levee protected area. The committee recommends: -43-
FEMA should contract for the development of a list of key categories concerning the physical condition of a levee that would be used to evaluate the levee's ability to function effectively and concerning use of those factors to estimate geotechnical risk. An unsatisfactory rating would result in increased insurance premiums. -44-
TABLE 2 EXAMPLE LEVEE RATING PROCEDURE 1. Indicators are estimated on a 3-point scale (3 grade levels) for factors grouped by categories. Criteria for each grade level are explicitly detailed. An inspector visits site and reviews the physical condition of the levee and related facilities, and following the criteria, completes the check list detailed in Table 3. Each factor in the check list is then rated as unsatisfactory, work needed, or satisfactory, and a score of 0, l, or 2 is awarded as indicated in Table 3. 2. Grading sheet data are combined through a specified formula to assign a grade (G) for each category. 3. The grades (G) for each of the categories are combined in a weighted overall score where W^ is the weight given to the first category and G^ is the actual numerical value for the first category, for example. 4. Depending on the score ( S) , property owners will either receive, or be denied, a reduction in insurance rates. This reduction would be deducted from the premium paid by a property owner behind a levee that has an adequate rating. -45-
TABLE 3 RATING CHECK LIST: AN EXAMPLE Grading Factors I. Levee Design (Physical Features) 1.1 Homogeneous cross-section 1.2 Absence of settlement 1.3 Etc. II. Maintenance Program 11.1 Absence of erosion 11.2 Absence of animal burrows 11.3 Etc. III. Closures 111.1 Functional operating mechanisms 111.2 Etc. IV. Etc. Unsatis- Work factory needed 0 1 Satisfactory 2 Note: Specific ratings by factor are marked by checking as appropriate. -46-
