SQRA Calculation Methodology

RMC-TN-2018-01

1E-01
Annual Probability of Failure

1E-02
~, PFM

1E-03
' ', 1
Total

I'' ,, PFM

1E-04
' 2

I'' ,,
'
1E-05
', ,,
1E-06 '
1 10 100 1,000
Average Life Loss

US Army Corps
of Engineers ®
Institute for Water Resources
Risk Management Center
 Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188

The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data
sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other
aspect of this collection of information, including suggestions for reducing the burden, to Department of Defense, Washington Headquarters Services, Directorate for Information
Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other
provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number.
PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From - To)
04-17-2018
4. TITLE AND SUBTITLE
I Technical Note
5a. CONTRACT NUMBER
Semi-Quantitative Risk Analysis Calculation Methodology
5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

Timothy M. O’Leary, P.E.
5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION

Risk Management Center REPORT NUMBER

12596 West Bayaud Ave. Suite 400 RMC-TN-2018-01

Lakewood, CO 80228
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR'S ACRONYM(S)

11. SPONSOR/MONITOR'S REPORT

NUMBER(S)
RMC-TN-2017-06
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution is unlimited.

13. SUPPLEMENTARY NOTES

14. ABSTRACT
This document summarizes the methodology to estimate total incremental risk, non-breach risk for dams, and
overtopping incremental and non-breach risk for levees for semi-quantitative risk analysis performed by the U.S.
Army Corps of Engineers.

15. SUBJECT TERMS

Semi-quantitative risk analysis, risk assessment, dam safety, levee safety, overtopping

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF RESPONSIBLE PERSON
ABSTRACT OF
a. REPORT b. ABSTRACT c. THIS PAGE
PAGES
Timothy M. O’Leary
U U U UU 14 19b. TELEPHONE NUMBER (Include area code)
502-315-6599
Standard Form 298 (Rev. 8/98)
Prescribed by ANSI Std. Z39.18
SQRA Calculation Methodology

April 2018

U.S. Army Corps of Engineers

Institute for Water Resources
Risk Management Center
12596 West Bayaud Ave. Suite 400
Lakewood, CO 80228
 SQRA Calculation Methodology

Author
The results, findings, and recommendations provided in this document are technically sound and
consistent with current Corps of Engineers practice.

Timothy M. O’Leary, P.E.

Senior Geotechnical Engineer
Risk Management Center

Reviewed
This report has been checked and reviewed and is believed to be in accordance with the
standards of the profession.

David A. Margo, P.E.

Lead Civil Engineer
Risk Management Center

Approved

Nathan J. Snorteland, P.E.

Director
Risk Management Center
 SQRA Calculation Methodology

Contents
Executive Summary ......................................................................................................................... 1
Risks Posed by Dams and Levees .................................................................................................. 1
Overtopping Risk Posed by Levees ................................................................................................. 3
Non-Breach Risk ........................................................................................................................... 11
Levees ....................................................................................................................................... 12
Dams.......................................................................................................................................... 15

Figures
Figure 1 - Incremental Life Safety Risk Matrix for SQRA ................................................................. 2
Figure 2 - Incremental Economic Risk Matrix for SQRA .................................................................. 2
Figure 3 - Incremental Life Safety Risk Matrix Example .................................................................. 3
Figure 4 - Example Societal Risk Estimate based on Critical Overtopping Elevation ...................... 5
Figure 5 - Example Societal Risk Estimate Based on a Range of Overtopping Events ................... 7
Figure 6 - Divergent Consequence Example for Levees ................................................................. 8
Figure 7 - Incremental Risk Example for a Levee with Convergent Life Loss .................................. 9
Figure 8 - Incremental Risk Example for a Levee with Divergent Life Loss ................................... 11
Figure 9 - Non-Breach Risk Matrix for SQRA ................................................................................ 12
Figure 10 - Non-Breach Economic Risk Matrix for SQRA.............................................................. 12
Figure 11 - Non-Breach Risk Example for Levees ......................................................................... 14
Figure 12 - Refined Non-Breach Life Loss Relationship ................................................................ 15
Figure 13 - Non-Breach Life Safety Risk Matrix Example for Dams .............................................. 16
 SQRA Calculation Methodology

This page blank intentionally

 SQRA Calculation Methodology

Executive Summary
Each risk assessment includes estimates of the risk posed by the dam or levee (incremental risk),
the non-breach risk, and the flood risk (residual risk). The total risk posed by the dam or levee
from all potential failure modes is evaluated against the tolerable risk guideline for societal and
individual life safety risk. The Dam and Levee Senior Oversight Groups use estimates of total dam
and levee risk to classify each structure and recommend appropriate courses of action with
respect to the structure and the population at risk. The non-breach and flood risk are assessed,
considered, and communicated to provide more complete, transparent, and informed decisions.
This technical note describes the preferred methodology to calculate and portray the total risk
posed by a dam or levee, the risk posed by an overtopping potential failure mode, and the non-
breach risk. Risk analysts should use these methods to calculate and portray risks for Semi-
Quantitative Risk Analysis (SQRA). The concepts and methods can also be applied to
Quantitative Risk Analysis (QRA) when appropriate.

Risks Posed by Dams and Levees

During SQRA, order-of-magnitude estimates are made for both likelihood of failure and
consequences for individual risk-driving potential failure modes (PFM) based on existing data and
the available consequence estimates. The likelihood of failure is an estimate of the annual
probability of failure (APF) based on the strength of the evidence. The consequences are based
on a series of standard inundation scenarios developed by the USACE Modeling, Mapping, and
Consequence Production Center (MMCC). Judgment is applied for each PFM, and subjective
adjustments are made as needed. The incremental life safety risk matrix shown in Figure 1 is
similar to the societal risk chart used for quantitative risk assessment, with annual failure
probability on the vertical axis and average life loss on the horizontal axis. The product of the
annual failure probability and the average life loss is the average annual life loss (AALL). The
dashed red lines on the matrix represent the societal tolerable risk limit for AALL of 1E-03 lives per
year and the individual risk limit of 1E-04 per year. For SQRA, individual risk is approximated by
using the total APF for the structure.
The incremental economic risk matrix shown in Figure 2 is developed for each structure and is
used more frequently for low-head navigation projects and levee systems where economic
consequences may be a primary risk-driver. The vertical axis is annual failure probability, and the
horizontal axis is average economic loss. There are no tolerable risk limits for economic risk, and
the economic consequences are not compared to the life loss ranges to arrive at a value for
human life.
Cells correspond to order-of-magnitude divisions (e.g., APF between 1E-03 and 1E-04,
incremental life loss between 10 and 100 lives, and incremental economic loss between $10M and
$100M). The PFMs are plotted on the risk matrices as boxes of the same size as the grid to
represent the precision of the estimates made during SQRA. Estimates are portrayed to the
nearest half-order of magnitude.

1
 SQRA Calculation Methodology

1E-01 1E-01
Annual Probability of Failure

Annual Probability of Failure

1E-02
I'' 1E-02

' ',
1E-03
', ,, 1E-03

'
1E-04
', 1E-04

1E-05
' ', 1E-05
I'' ,,
1E-06 ' 1E-06

1 10 100 1,000 $10M $100M $1B $10B

Average Life Loss Average Economic Loss
Figure 1 - Incremental Life Safety Risk Matrix for SQRA Figure 2 - Incremental Economic Risk Matrix for SQRA

The total failure probability and average consequences are calculated from individual PFM
estimates using the center of the order-of-magnitude box as a point estimate. Because risk
estimates are portrayed and evaluated in logarithmic space, a geometric mean is used to
calculate the point estimate at the center of a box. The total APF is calculated by summing the
APF point estimates for all of the individual PFMs. This calculation assumes the PFMs are
mutually exclusive.
APFtotal = ∑APFi
The total AALL is calculated by summing the AALL for all of the individual PFMs. Recall that AALL
is equal to the product of the failure probability and the average consequences.
AALLtotal = ∑AALLi = ∑(APFi x Ni)
The total average consequence estimate is calculated as the weighted average of the
consequences for all of the individual PFMs.
N = ∑(APFi x Ni) / ∑APFi = AALLtotal / APFtotal
An example is provided below.
Total Risk Posed by a Dam or Levee
Given:
• PFM 1: APF = SQRT(1E-02 * 1E-03) = 3E-03 and N = SQRT(1 * 10) = 3
• PFM 2: APF = 3E-04 and N = 300

2
 SQRA Calculation Methodology

Solution:
• Total Individual Risk = Total APF = 3E-03 + 3E-04 = 3.3E-03
• Total Societal Risk = (3E-03)(3) + (3E-04)(300) = 9.9E-02
• N = 9.9E-02 / 3.3E-03 = 30
• APF: In log space, 3.3E-03 is closest to 3E-03. The total risk is portrayed as 1E-03 to
1E-02.
• Average Life Loss: In log space, 30 is the geometric mean of 10 and 100. The total average
life loss is portrayed as 10 to 100.
The risk estimate for this example is portrayed in Figure 3.

1E-01
Annual Probability of Failure

1E-02
~, PFM

1E-03
' ', ~, 1
Total

' ,,
PFM
2
1E-04
', ,,
'
1E-05
', ,,
1E-06 '
1 10 100 1,000
Average Life Loss

Figure 3 - Incremental Life Safety Risk Matrix Example

Overtopping Risk Posed by Levees

USACE defines the levee control location (LCL) as the point along the levee system that is likely
to overtop first based on the levee grade and water surface profiles. The maximum (most
frequent) annual chance exceedance (ACE) for levee overtopping is defined as ACETOL. In QRA,
the APF for an overtopping PFM is estimated by combining the loading (probability of overtopping)
and performance (conditional probability of breach given overtopping) over a range of overtopping
events. For SQRA, a single critical overtopping elevation that is likely to result in a breach is
typically used for a simplified and sufficient estimate of APF for an overtopping PFM. The APF
estimate is typically set equal to the flood frequency at the critical overtopping depth and is
defined as ACEOTB. The estimate of average consequences and societal risk also depends on the
consequences associated with overtopping floods of varying magnitude. For SQRA, a single
critical overtopping elevation is typically not sufficient to estimate consequences and societal risk.
Changes in consequences over a range of overtopping events needs to be considered in order to

3
 SQRA Calculation Methodology

estimate the overtopping risk posed by the levee. Estimating overtopping societal risk based on a
single critical elevation typically results in overestimation of the societal risk posed by the levee
and underestimation of the flood risk. When consequence information is available, overtopping
risk estimates need to consider how consequences change over a range of overtopping flood
events.
As an example, consider the information provided in Table 1. The levee overtops at a frequency of
about 1 in 250, and the critical overtopping depth is judged to be 2 feet of overtopping at a
frequency of about 1 in 280. For SQRA, the critical overtopping elevation approach assumes that
the levee will not breach below this level and will breach above this level. This is reflected in the
probability of breach values in the third column of Table 1. This is a reasonable simplifying
approximation for the purposes of an SQRA. At the critical overtopping depth, a life loss of 100 is
expected to occur if the levee does not breach. If the levee breaches, a life loss of 800 is
estimated resulting in an incremental life loss of 700 (800 – 100). Using a single point estimate for
the SQRA would result in an APF estimate of 3.5E-03, life loss of 700, and a societal risk estimate
of 2.5 (3.5E-03 * 700). This risk estimate is portrayed in Figure 4.
Table 1 - Flood scenarios with and without breach
Annual
Chance Probability of Life loss if the Life loss if the
Overtopping Incremental
Exceedance Breach Given Levee does Levee does
Depth (feet) Life Loss
of a Flood OT Depth not Breach Breach
Scenario
0.0038
1 0 0 0 0
(ACETOL)
0.0035
2 1.0 100 800 700
(ACEOTB)
0.0030 3 1.0 800 1,600 800
0.0025 6 1.0 2,000 2,400 400
0.0020 10 1.0 3,100 3,100 0

4
 SQRA Calculation Methodology

1E-01

Annual Probability of Failure

1E-02
', OT
1E-03
' ', ~,
1E-04
' ', ~,
' ',
1E-05
', ,,
1E-06 '
1 10 100 1,000
Average Life Loss

Figure 4 - Example Societal Risk Estimate based on Critical Overtopping Elevation

Notice that the single critical-elevation-based risk estimate does not consider the decrease in
incremental consequences that occurs with increasing flood magnitude. This decrease in
incremental consequences commonly occurs with levees because the non-breach consequences
increase at a faster rate than the breach consequences. This behavior can be influenced by the
size of the leveed area, volume of the leveed area, rate of rise in the leveed area, depth of
overtopping, overtopping discharge, duration of overtopping, overtopping volume, breach
formation time, breach size, and shape of the flood hydrograph. Some questions to consider might
include the following:
• Is there sufficient overtopping volume to fill the leveed area with and without a levee
breach?
• Is there sufficient overtopping volume to fill the leveed area prior to reaching the critical
overtopping elevation?
• Does the extent and depth of flooding progress at a similar rate with and without a levee
breach?
• At what flood magnitude does the levee become overwhelmed?
• Do consequences continue to increase with increasing flood magnitude, or do they begin to
taper off?
Using the available consequence information in Table 1, the estimate of the overtopping risk
posed by the levee can be improved. A societal risk estimate is obtained by combining a loading
function, a system response function, and a consequence function and then integrating to
estimate the average annual life loss. The preferred calculation method shown in Table 2 is a
discrete approximation of the integration. This is the same type of calculation used in event tree
analysis.

5
 SQRA Calculation Methodology
Table 2 – Calculating Risk Posed by the Levee for an Overtopping Potential Failure Mode
Annual
Probability
Chance Average
of Breach Incremental
Exceedance Calculation Annual
Given OT Life Loss
of a Flood Life Loss
Depth
Scenario
0.0038
0 0 (0.0038 – 0.0035) * 0 * (0 + 700)/2 0
(ACETOL)
0.0035
1.0 700 (0.0035 – 0.003) * 1 * (700 + 800)/2 0.38
(ACEOTB)
0.0030 1.0 800 (0.003 – 0.0025) * 1 * (800 + 400)/2 0.30
0.0025 1.0 400 (0.0025 – 0.002) * 1 * (400 + 0)/2 0.1
0.0020 1.0 0 0.002 * 1 * 0 0

The estimate of APF is the same as the previous estimate of 3.5E-03. The estimate of AALL is
based on the product of the flood interval probability, the system response, and the average
consequences over the flood interval. In this calculation, the system response is implicitly
assumed to be a step function. The probability of breach is 0 below the critical overtopping depth
and 1 above the critical overtopping depth. This is typically a reasonable and sufficient assumption
for SQRA. The total AALL is calculated by summing the values in the last column of Table 1.
AALL = 0 + 0.38 + 0.30 + 0.1 + 0 = 0.8
The average life loss is calculated by dividing the AALL by the APF.
N = 0.8 / 3.5E-03 = 230
The improved risk estimate is portrayed in Figure 5. Notice that the revised risk estimate is one
order of magnitude less than the estimate based on a single critical overtopping elevation. This
demonstrates the impact consequences can have on overtopping risk estimates. When
consequence information is available or can be estimated by judgment, this is the preferred
approach for estimating and portraying overtopping risk for SQRA. The approach can also be
used for QRA. Refinements can be made to the system response and other inputs as needed to
achieve the desired precision and accuracy needed for the risk estimate.

6
 SQRA Calculation Methodology

1E-01

Annual Probability of Failure

1E-02
', OT
1E-03
' ', ~,
1E-04
' ', ~,
' ',
1E-05
', ,,
1E-06 '
1 10 100 1,000
Average Life Loss

Figure 5 - Example Societal Risk Estimate Based on a Range of Overtopping Events

As shown in the example, the consequences with and without a levee breach often converge at
some flood event such that the incremental consequences become zero. In some cases,
interpolation or extrapolation of consequence information is necessary to estimate the ACE for this
flood event. This is the flood where the consequences with or without breach of the levee are
equivalent. This is an important point for the risk estimate because the risk posed by the levee
(incremental risk) is zero for floods less frequent than this point.
Another important event is the flood that has the highest incremental consequences. As shown in
the example, this flood event may not be the same as the critical overtopping event.
The consequences with and without a levee breach do not converge for every levee. The
incremental consequences could continue to increase with increasing flood magnitude. In these
cases, it is necessary to estimate an overtopping flood that captures the majority of the societal
risk. This flood should produce widespread flooding of the leveed area such that larger floods
would not substantially change the average consequence estimate. This flood should also be
infrequent enough, but still plausible, such that the probability contribution from larger floods would
not substantially change the risk estimate. USACE defines this overtopping event as the ultimate
overtopping flood (ACEult). For SQRA, the incremental life loss is assumed to be a constant value
for flood events less frequent than ACEult. The concept is illustrated in Figure 6.

7
 SQRA Calculation Methodology

Jtl J _ _ J _ J __ 1 ____
900
800 -- -- Breach
-- Non-Breach ,,
-- - - -Incremental
L"
Average Life Loss (lives)
700
600
500 ~
,r
/'
~
400
300 /
200
100
0
0.001 ACEOT ACEOTB 1E-4 ACEult 1E-5
Annual Chance Exceedance, ACE (floods/year)

Figure 6 - Divergent Consequence Example for Levees

Examples for a convergent (zero incremental consequences) and divergent (increasing
incremental consequences) scenario are provided below. The examples and calculations assume
linear relationships to calculate the societal risk. Depending on the form of the relationships,
greater precision could be obtained by discretizing more intervals between the given points and
performing interpolation using alternative scales such as a probability scale (z-variate) or a
logarithmic scale. The selected scale should be consistent with the expected shape of the
relationship to provide a more accurate interpolation.
Convergent Example for Levees
Given:
• Frequency of overtopping event to initiate breach, ACEOTB = 3.51-03
• Breach and non-breach life loss:
OT Depth Breach Life Loss
Overtopping Event ACE
(ft) Day Night Average
Top of levee (TOL) 0.0 3.70E-03 596 897 777
1-ft overtopping (OTB) 1.0 3.51E-03 596 897 777
1.5-ft overtopping 1.5 3.33E-03 870 1,373 1,172
6.0-ft overtopping 6.0 2.00E-03 2,347 3,635 3,120

OT Depth Non-Breach Life Loss

Overtopping Event ACE
(ft) Day Night Average
Top of levee (TOL) 0.0 3.70E-03 0 0 0
1-ft overtopping (OTB) 1.0 3.51E-03 3 8 6
1.5-ft overtopping 1.5 3.33E-03 173 297 247
6.0-ft overtopping 6.0 2.00E-03 2,343 3,633 3,117

8
 SQRA Calculation Methodology

Solution:
• Average incremental life loss = average breach life loss – average non-breach life loss
OT Depth Incremental Life Loss
Overtopping Event ACE
(ft) Day Night Average
Top of levee (TOL) 0.0 3.70E-03 596 897 777
1-ft overtopping (OTB) 1.0 3.51E-03 593 889 771
1.5-ft overtopping 1.5 3.33E-03 697 1,076 924
6.0-ft overtopping 6.0 2.00E-03 4 2 3

• Using judgment and extrapolation, incremental consequences are estimated to be zero at

an ACE of 1.997E-03 and an overtopping depth of 6.1 feet.
• APFOT = ACEOTB = 3.51E-03
• AALLOT = (3.51E-03 – 3.33E-03)(771 + 924)/2 + (3.33E-03 – 2.00E-03)(924 + 3)/2
+ (2.00E-03 – 1.997E-03)(3 + 0)/2 = 7.69E-01
• N = (7.69E-01) / (3.51E-03) = 219
• Incremental Life Safety Risk Matrix:
o Annual Probability of Failure: between 1E-03 and 1E-02
o Average Life Loss: between 100 and 1,000
The risk estimate for this example is portrayed in Figure 7.

1E-01
Annual Probability of Failure

1E-02
~,
OT
1E-03
' ', ~,
' ,,
1E-04
', ,,
'
1E-05
', ,,
1E-06 '
1 10 100 1,000
Average Life Loss

Figure 7 - Incremental Risk Example for a Levee with Convergent Life Loss

9
 SQRA Calculation Methodology

Divergent Example for Levees

Given:
• Frequency of overtopping event to initiate breach, ACEOTB = 1.60-04
• Breach and non-breach life loss:
OT Depth Breach Life Loss
Overtopping Event ACE
(ft) Day Night Average
Top of levee (TOL) 0.0 6.59E-04 216 331 285
1.0-ft overtopping 0.5 2.89E-04 216 331 285
1.4-ft overtopping (OTB) 1.4 1.60E-04 427 615 540
2.0-ft overtopping 2.0 7.14E-05 702 986 872

OT Depth Non-Breach Life Loss

Overtopping Event ACE
(ft) Day Night Average
Top of levee (TOL) 0.0 6.59E-04 0 0 0
1.0-ft overtopping 0.5 2.89E-04 0 0 0
1.4-ft overtopping (OTB) 1.4 1.60E-04 7 13 11
2.0-ft overtopping 2.0 7.14E-05 17 31 25

Solution:
• Average incremental life loss = average breach life loss – average non-breach life loss
OT Depth Incremental Life Loss
Overtopping Event ACE
(ft) Day Night Average
Top of levee (TOL) 0.0 6.59E-04 216 331 285
1.0-ft overtopping 0.5 2.89E-04 216 331 285
1.4-ft overtopping (OTB) 1.4 1.60E-04 420 602 529
2.0-ft overtopping 2.0 7.14E-05 685 955 847

• Using judgment, 2 feet of overtopping is judged to be adequate to capture most of the

societal risk. This event is selected as the ultimate overtopping.
• Assume the incremental life loss for 2.0-ft OT is constant for less frequent events.
• APFOT = ACEOTB = 1.60E-04
• AALLOT = (1.60E-04 – 7.14E-05)(529 + 847)/2 + (7.14E-05)(847) = 1.21E-01
• N = (1.21E-01) / (1.60E-04) = 756
• Incremental Life Safety Risk Matrix:
o Annual Probability of Failure (APF): between 3E-05 and 3E-04
o Average Life Loss: between 300 and 3,000
The risk estimate for this example is portrayed in Figure 8.

10
 SQRA Calculation Methodology

1E-01

Annual Probability of Failure

1E-02
I''
1E-03
' ', ~,
' ', OT
1E-04
',
' ',
1E-05
', ,,
1E-06 '
1 10 100 1,000
Average Life Loss

Figure 8 - Incremental Risk Example for a Levee with Divergent Life Loss

Non-Breach Risk
Dams and levees use a consistent approach for estimating non-breach risk. The ACE when the
public would begin to experience flooding due to levee overtopping or dam spillway release and
the ACE when life loss would start to occur are important to communicate flood risk to the public.
For levees, the ACE for flooding is typically when the levee begins to overtop. For dams, the ACE
for flooding is typically at the top of active storage pool elevation. The ACE when life loss would
start to occur depends on the specific situation but is typically less than the ACE for flooding.
Example matrices for portraying the non-breach risk are shown in Figure 9 and Figure 10.

11
 SQRA Calculation Methodology

1E-01 1E-01

Annual Probability of Economic Loss

Annual Probability of Life Loss

1E-02 1E-02

1E-03 1E-03

1E-04 1E-04

1E-05 1E-05

1E-06 1E-06

1 10 100 1,000 $10M $100M $1B $10B

Average Life Loss Average Economic Loss
Figure 9 - Non-Breach Risk Matrix for SQRA Figure 10 - Non-Breach Economic Risk Matrix for SQRA

Levees
USACE does not typically consider interior flooding scenarios where water flowing into the leveed
area from the landside of the levee exceeds the available pumping capacity; therefore, the ACE
associated with the top of levee is used to estimate the likelihood of flooding. The annual
probability of inundation due to overtopping without breach (APINB) is estimated as the ACE at the
top of the levee. The estimated ACE when life loss begins to occur (≥ 1) depends on the available
consequence information. For levees, it is likely that the ACE for overtopping and the ACE for life
loss are similar enough to be considered equivalent. However, this is not necessarily true for
every levee.
The estimation of non-breach risk assumes that consequences are constant for flood events less
frequent than the ultimate overtopping flood. Recall that this event is selected so that most of the
risk is captured in the estimate. This flood event is typically greater in magnitude (less frequent)
than the flood event for initial overtopping and greater than the flood event for the critical
overtopping depth. Releases from upstream dams can also have a significant effect on the non-
breach risk estimate for the leveed area. Failure to consider these larger flood events results in an
underestimation of the non-breach and flood risk. It is important that the ultimate overtopping flood
reflect a plausible maximum scenario. Available consequence information and judgment should be
used to select an ultimate overtopping load for non-breach risk estimates.

12
 SQRA Calculation Methodology

Non-Breach Example for Levees

Given:
• Non-breach life loss:
OT Depth Non-Breach Life Loss
Overtopping Event ACE
(ft) Day Night Average
Top of levee (TOL) 0.0 3.70E-03 0 0 0
1.0-ft overtopping 1.0 3.51E-03 3 8 6
1.5-ft overtopping 1.5 3.33E-03 173 297 247
6.0-ft overtopping 6.0 2.00E-03 2,343 3,633 3,117

Solution:
• In the previous example, a slight extrapolation was performed to estimate the flood with
zero incremental life loss. This estimate is used in this example for the ultimate overtopping
event. The ACE is 1.997E-3 and the life loss is 3,126.
OT Depth Non-Breach Life Loss

-----
Overtopping Event ACE
(ft) Day Night Average
Ultimate overtopping flood 6.01 1.997E-03 3,126

• APINB = ACETOL = 3.70E-03

• ACE for life loss initiation = 3.70E-3
• AALLNB = (3.70E-03 – 3.51E-03)(0 + 6)/2 + (3.51E-03 – 3.33E-03)(6 + 247)/2
+ (3.33E-03 – 2.00E-03)(247 + 3,117)/2 + (2.00E-03 – 1.997E-03)(3,117 + 3,126)/2
+ (1.997E-03)(3,126) = 8.51
• NNB = 8.51 / 3.70E-03 = 2,300
• Non-Breach Life Safety Risk Matrix:
o Likelihood of Flooding: between 1E-03 and 1E-02
o Average Life Loss: between 1,000 and 10,000
The risk estimate for this example is portrayed in Figure 11.

13
 SQRA Calculation Methodology

1E-01

Annual Probability of Life Loss

1E-02

NB
1E-03 Levee
Overtops
1E-04 at 1 in 270

1E-05

1E-06

1 10 100 1,000
Average Life Loss

Figure 11 - Non-Breach Risk Example for Levees

14
 SQRA Calculation Methodology

Dams
The standard inundation scenarios for USACE include Top of Active Storage (TAS) and Maximum
High Pool (MHP) elevations. The MHP elevation corresponds to the Probable Maximum Flood
(PMF) for high hazard potential projects. For an uncontrolled spillway, TAS would typically be the
elevation of the spillway crest. For a gated spillway, this would typically be an elevation near the
top of the spillway gates. This elevation is sometimes referred to as the “top of flood control pool.”
The water control manual and water management staff should be consulted for the official TAS
elevation.
The methodology for estimating non-breach risk for dams is similar to levees. The ACE for
flooding is defined by the top of active storage which is when surcharge operations (spillway
releases) would typically initiate. The ACE when life loss begins to occur is estimated based on
the available consequence information and judgment. For dams, it is likely that the ACE for life
loss is less than the ACE for flooding. However, this is not necessarily true for every dam. An
assessment of the dam outflow required to pose a threat to life safety or to cause extensive
flooding or damage can inform the ACE estimate. Moderate or major flood stages published by
the National Weather Service (http://water.weather.gov/ahps/) may provide additional insights.
The concepts are illustrated in Figure 12 using available consequence information to estimate the
ACE for initiation of life loss. In this example, the ACE of flooding is about 1 in 200 and the ACE
for initiation of life loss is about 1 in 1,000.
TAS N>0 MHP
40

35
Average Life Loss (lives)

30
/f -
25 ,,,,
20
~ . ~
V
.... ~
15
..... ~~ -
/.
_, 1~" -
~'I
10
V
/
5 ,..,, . ,,, ~ -
~

0
0.01 0.001 1E-4 1E-5
Annual Chance Exceedance, ACE (floods/year)

Figure 12 - Refined Non-Breach Life Loss Relationship

The annual probability of non-breach inundation (API) is estimated as the ACE at top of active
storage. The annual probability of consequences is estimated as ACEN>0 for life safety risk and
ACE$>0 for economic risk.
An example is provided below. This example and the calculations assume a linear relationship
between the given points. Additional points and non-linear interpolation methods could be used to
increase the precision and accuracy of the estimate as needed. The level of detail in this example
is sufficient for SQRA.

15
 SQRA Calculation Methodology

Non-Breach Example for Dams

Given:
• Consequence scenarios for TAS and MHP.
• Existing data is used to assess the dam outflow corresponding to life-threatening releases
at 1.00E-03 ACE flood event.
• Non-breach life loss:
Non-Breach Life Loss
Overtopping Event ACE
Day Night Average
Top of Active Storage 5.00E-03 0 0 0
Threshold for non-zero life loss 1.00E-03 0 0 0
Maximum High Pool 1.00E-04 27 32 30

Solution:
• APINB = ACEN>0 = 1.00E-03
• ACE for flooding is 5E-03
• AALLNB = (1.00E-03 – 1.00E-04)(0 + 30)/2 + (1.00E-04)(30) = 1.65E-02
• NNB = (1.65E-02 lives) / (1.00E-03) = 17
• Non-Breach Life Safety Risk Matrix:
o Likelihood of Flooding: between 3E-04 and 3E-03
o Average Life Loss: between 3 and 30
The risk estimate for this example is portrayed in Figure 13.

1E-01
Annual Probability of Life Loss

1E-02

Top of
1E-03 NB
Active
1E-04
Storage
is 1 in 200
1E-05

1E-06

1 10 100 1,000
Average Life Loss

Figure 13 - Non-Breach Life Safety Risk Matrix Example for Dams

16