From 4b6bc9cbd2dcf862a14dc0f73fc477d2cf655df2 Mon Sep 17 00:00:00 2001 From: Adam Gohs Date: Thu, 15 Oct 2026 18:31:45 -0400 Subject: [PATCH 1/2] Add the Concentrated Leak Erosion Common Methods Screening technical reference manual to the System Response Library New document, registered as an active draft. --- .../v1.0/00-document-info.mdx | 25 ++ .../v1.0/00-version-history.mdx | 19 ++ .../v1.0/01-preface.mdx | 50 +++ .../v1.0/02-model-overview.mdx | 193 +++++++++++ .../v1.0/03-input-parameters.mdx | 312 ++++++++++++++++++ .../v1.0/04-event-tree.mdx | 159 +++++++++ .../v1.0/05-flaw-conventions.mdx | 210 ++++++++++++ .../v1.0/06-initiation.mdx | 234 +++++++++++++ .../07-failure-pathway-and-progression.mdx | 205 ++++++++++++ .../v1.0/08-assumptions-and-limitations.mdx | 159 +++++++++ .../v1.0/09-output-and-interpretation.mdx | 130 ++++++++ .../v1.0/_input-needed.md | 249 ++++++++++++++ src/docConfig.js | 2 +- .../v1.0/bib.json | 17 + 14 files changed, 1963 insertions(+), 1 deletion(-) create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/00-document-info.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/00-version-history.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/01-preface.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/02-model-overview.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/03-input-parameters.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/04-event-tree.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/05-flaw-conventions.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/06-initiation.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/07-failure-pathway-and-progression.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/08-assumptions-and-limitations.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/09-output-and-interpretation.mdx create mode 100644 docs/web-applications/system-response/cle-common-methods-screening/v1.0/_input-needed.md create mode 100644 static/bibliographies/web-applications/system-response/cle-common-methods-screening/v1.0/bib.json diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/00-document-info.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/00-document-info.mdx new file mode 100644 index 000000000..c55c5b078 --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/00-document-info.mdx @@ -0,0 +1,25 @@ +--- +title: Document Info +reportDate: October 2026 +reportType: Technical Reference +reportTitle: Concentrated Leak Erosion Common Methods Screening Technical Reference Manual +reportSubTitle: System Response Library +reportAuthors: ['Adam Gohs, P.E.'] +reportAbstract: Technical reference documenting the methods shared by the eight concentrated leak erosion (CLE) screening models in the System Response library. Each model computes a system response probability (SRP) at each headwater level from an eight-node event tree that runs from a flaw through initiation, continuation, three progression conditions, intervention, and breach. This manual documents what the models have in common, namely the event-tree structure and node conventions, the risk-factor and likelihood-factor scoring used to estimate flaw probabilities, the two initiation treatments (a headwater gate at the bottom of an open crest crack, and a cylindrical-pipe hydraulic shear stress analysis for compaction flaws, with its shared critical shear stress table), the failure-pathway conventions that gate continuation, roof support, and flow limitation, and the nodal results the models report. The mechanism-specific flaw assessments are documented in each model's own manual. +reportSubjectTerms: ['CLE', 'Concentrated Leak Erosion', 'Screening', 'Event Tree', 'System Response Probability', 'System Response Curve', 'Risk Factor', 'Likelihood Factor', 'Critical Shear Stress', 'Cylindrical Pipe', 'Roof Support', 'Crack Filling', 'Flow Limitation', 'Monte Carlo'] +citationGuide: 'U.S. Army Corps of Engineers, Risk Management Center, Concentrated Leak Erosion Common Methods Screening Technical Reference Manual, Lakewood, CO: U.S. Army Corps of Engineers, Risk Management Center, 2026.' +--- + +import Link from '@docusaurus/Link'; +import DocumentMetadata from '@site/src/components/DocumentMetadata'; +import NavContainer from '@site/src/components/NavContainer'; + + + +# Document Information + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/00-version-history.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/00-version-history.mdx new file mode 100644 index 000000000..fbec3b621 --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/00-version-history.mdx @@ -0,0 +1,19 @@ +--- +title: Version History +--- + +import NavContainer from '@site/src/components/NavContainer'; +import TableVersionHistory from '@site/src/components/TableVersionHistory'; + + + +# Version History + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/01-preface.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/01-preface.mdx new file mode 100644 index 000000000..eb83a142d --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/01-preface.mdx @@ -0,0 +1,50 @@ +--- +title: Preface +--- + +import NavContainer from '@site/src/components/NavContainer'; + + + +# Preface + +This document provides a technical reference for the methods shared by the eight concentrated leak erosion (CLE) screening models in the System +Response library. Each model computes a system response probability (SRP) as a function of headwater elevation for one CLE mechanism: cross-valley +differential settlement, differential settlement adjacent to a wall or cliff, poor compaction within the core, poor compaction adjacent to a conduit, +poor compaction adjacent to a wall or cliff, poor compaction along the foundation or abutment contact, desiccation cracking at the crest, and frost +cracking at the crest. The models are screening-level: they apply simplifications intended to reduce input burden while avoiding false negatives, and +any application that performs screening-level risk analysis can use them. + +The eight models differ in how they estimate the probability that a flaw exists and in where that flaw sits in the embankment. Everything else is +common to the family: the eight-node event tree and how each node is populated, the scoring scheme used to turn qualitative site factors into a flaw +probability, the two initiation treatments, the failure-pathway rules that decide which filter, materials, and upstream elements apply at each pool +level, and the form of the results. This manual documents those common methods once. Each model's own technical reference manual documents its flaw +assessment and its mechanism-specific inputs, and refers to this manual for the rest. The continuation (filter evaluation) node is evaluated by a +shared continuation method with its own manual, the Continuation (Filter Evaluation) Screening Technical Reference Manual; this manual documents only +when that method is applied. + +The intended audience includes dam safety engineers evaluating whether the CLE screening models are appropriate for their assessment, and peer +reviewers checking the technical basis of screening-level results. Readers are assumed to have a general background in internal-erosion risk +analysis — event trees, concentrated leak erosion mechanics, and probability concepts — but no familiarity with the software implementation is +required. + +## Document Organization + +- **[Model Overview](./02-model-overview.mdx)** — The purpose and intended use of the CLE screening models, a summary of their common execution + pathway, their engineering basis, a comparison with detailed methods, the two groups the eight models fall into, and their relationship to the + toolbox methods. +- **[Common Input Parameters](./03-input-parameters.mdx)** — The inputs shared across the family: hydraulic loading, the units and datum + convention, the embankment description, the failure-pathway inputs, the upstream seepage-limiting elements, the initiation material inputs of the + compaction models, and the intervention and breach probabilities. +- **[Event Tree](./04-event-tree.mdx)** — The eight event-tree nodes, which nodes vary with pool level, how the headwater grid is formed, and how the + SRP is computed. +- **[Flaw Conventions](./05-flaw-conventions.mdx)** — The risk-factor and likelihood-factor scoring scheme, the optional flaw-continuity stage, + observation overrides, and the flaw elevation each model passes to the downstream nodes. +- **[Initiation](./06-initiation.mdx)** — The headwater gate used by the crest-crack models, and the cylindrical-pipe initiation analysis used by + the compaction models, with its critical shear stress and initial pipe diameter defaults. +- **[Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx)** — How the failure pathway and any intersecting filter gate the + continuation and roof-support nodes, and how the crack-filling and flow-limitation nodes are evaluated. +- **[Assumptions and Limitations](./08-assumptions-and-limitations.mdx)** — Each common modeling assumption with its engineering basis, its effect + on results, and the conditions under which it may not hold. +- **[Output and Interpretation](./09-output-and-interpretation.mdx)** — The results the models produce, how to read the system response curve and + its node breakdown, and the role of screening results relative to detailed analysis. diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/02-model-overview.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/02-model-overview.mdx new file mode 100644 index 000000000..87accfa5e --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/02-model-overview.mdx @@ -0,0 +1,193 @@ +--- +title: Model Overview +--- + +import Citation from '@site/src/components/Citation'; +import CitationFootnote from '@site/src/components/CitationFootnote'; +import NavContainer from '@site/src/components/NavContainer'; +import TableReference from '@site/src/components/TableReference'; +import TableVertical from '@site/src/components/TableVertical'; + + + +# Model Overview + +## Purpose + +Each CLE screening model computes a system response curve for one concentrated leak erosion mechanism as a function of headwater elevation. The +system response probability (SRP) at each headwater level is the conditional probability of breach of the embankment by concentrated leak erosion +through a flaw of that mechanism, given that pool level. The curve of SRP values across the range of headwater levels characterizes the +embankment's vulnerability to the mechanism. + +Concentrated leak erosion is erosion of the walls of an open crack, gap, or poorly compacted zone by water flowing through it. The failure mode +requires a continuous flaw through the embankment, a hydraulic shear stress on the flaw walls large enough to initiate erosion, an unfiltered exit +that lets the eroded soil leave the embankment, an enlarging pipe that stays open and is neither filled nor starved of flow, unsuccessful +detection and intervention, and finally an uncontrolled release of the impounded water. + +## Intended Use + +The CLE screening models are screening-level assessments of concentrated leak erosion risk. They make simplifying and generally conservative +assumptions with two goals: + +- **Preventing false negatives** — The models are designed to avoid underestimating risk at the screening level, so that embankments with potential + concentrated leak erosion concerns are flagged for further analysis. +- **Reducing input burden** — Qualitative selections (compaction categories, climate categories, conduit and wall descriptions, the embankment + type) substitute for site-specific measurements, and default distributions cover the erodibility and initial-opening properties that are + typically unavailable at the screening stage. + +The models are not intended for detailed or semi-quantitative risk assessments. For higher-level analyses, the full concentrated leak erosion +toolbox methods and the internal-erosion probability methods they draw on provide the complete methodologies with full parameterization (see +[Comparison with Detailed Models](#comparison-with-detailed-models)). + +## How the Model Works + +Every CLE screening model executes the same three stages: + +1. **Flaw assessment** — The model estimates the probability that a continuous flaw of its mechanism exists through the embankment, and the + elevation of that flaw. The probability comes from scoring qualitative and geometric site factors (or, for cross-valley settlement, from a + settlement-based relationship), and an observation of the flaw — leakage, wet spots, or observed continuous cracking, depending on the model — + can override it to certainty. The flaw probability is the same at every pool level (see [Flaw Conventions](./05-flaw-conventions.mdx)). + +2. **Pathway resolution** — The model resolves the failure pathway through the embankment at the flaw elevation: whether a filter intersects it + and where the filter's top lies, which materials must hold a roof over the enlarging pipe, and which upstream seepage-limiting elements still + reach the flaw (see [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx)). + +3. **Event-tree assembly** — At each headwater level, the model composes the eight node probabilities and multiplies them to produce the SRP. + Initiation is either a gate on the headwater rising above the bottom of an open crest crack, or a Monte Carlo cylindrical-pipe analysis of the + hydraulic shear stress at the flaw (see [Initiation](./06-initiation.mdx) and [Event Tree](./04-event-tree.mdx)). + +## Physics Basis + +The family combines the following engineering methods: + +- **Risk-factor and likelihood-factor scoring** for the probability of a continuous flaw, in which each site factor is assigned a likelihood + factor, weighted by a risk factor, and the weighted sum is mapped to a flaw probability (see + [Flaw Conventions](./05-flaw-conventions.mdx)) +- **Hydraulic shear stress on the walls of a cylindrical pipe** compared with the soil's critical shear stress for the initiation of erosion in a + compaction flaw, as documented in the + [Cylindrical Pipe](/docs/toolbox-technical-manuals/internal-erosion-suite/concentrated-leak-erosion-initiation/v1.0/cylindrical-pipe) chapter + of the Concentrated Leak Erosion Initiation manual +- **Roof-support probabilities by soil classification** for the ability of the enlarging pipe to stay open, and **crack-filling probabilities by + embankment type** for the chance that upstream material fills the pipe +- **Flow-limitation probabilities by upstream element** for the chance that a cutoff wall or upstream facing restricts the flow through the + flaw +- **Filter evaluation** for the probability that erosion continues past a filter intersecting the failure pathway, applied through the shared + continuation method documented in the Continuation (Filter Evaluation) Screening Technical Reference Manual and drawing on the + [Filter Evaluation (Continuation)](/docs/toolbox-technical-manuals/internal-erosion-suite/filter-evaluation-continuation/v1.0/general-overview) + toolbox methods + +## Comparison with Detailed Models + +The detailed counterparts of the screening models are the concentrated leak erosion toolbox methods and the internal-erosion probability methods +of Fell et al. (2024) applied in full. The Concentrated Leak Erosion Initiation manual covers four flaw +geometries — a cylindrical pipe, a horizontal crack, a vertical rectangular crack, and a vertical triangular crack (see its +[General Overview](/docs/toolbox-technical-manuals/internal-erosion-suite/concentrated-leak-erosion-initiation/v1.0/general-overview)) — of +which the compaction models use the cylindrical pipe. summarizes the differences. + + + +## The Two Model Groups + +The eight models fall into two groups that share their flaw geometry and therefore their initiation treatment and their pathway rules +(): + +- **Crest-crack models** — cross-valley differential settlement, differential settlement adjacent to a wall or cliff, desiccation cracking, and + frost cracking. The flaw is a transverse crack open at the crest and extending down to a bottom-of-crack elevation. Every node is + deterministic except continuation, which the shared continuation method may evaluate by Monte Carlo simulation when a filter intercepts the + pathway (its route through the USCS classes of the base and filter soils samples their gradation parameters). +- **Compaction models** — poor compaction within the core, adjacent to a conduit, adjacent to a wall or cliff, and along the foundation or + abutment contact. The flaw is a zone of poorly compacted soil at a fixed base-of-flaw elevation, idealized for initiation as a cylindrical pipe. + Initiation is evaluated by Monte Carlo simulation, and continuation may be as well, by the same continuation route as the crest-crack models; + every other node is deterministic. + + + +The flaw elevation is the one flaw quantity every downstream node uses: it sets where initiation can begin, where the failure pathway runs, and +which filter and upstream elements the pathway meets. [Flaw Conventions](./05-flaw-conventions.mdx#flaw-elevation) documents how each model +resolves it. + +## Relationship to Toolbox Models + +The compaction models use the same cylindrical-pipe initiation analysis as the standalone Concentrated Leak Erosion Initiation toolbox. They +differ mainly in what they supply to it: the embankment section and flaw elevation come from the screening inputs, and the critical shear stress and +initial pipe diameter default to screening ranges that the user may replace. [Initiation](./06-initiation.mdx#delegated-method-cylindrical-pipe-initiation) +documents each supplied quantity and its basis. The crest-crack models use no toolbox method for initiation. All eight models hand the continuation node to the +shared continuation method whenever a filter intersects the failure pathway +(see [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#node-3-continuation)). + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/03-input-parameters.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/03-input-parameters.mdx new file mode 100644 index 000000000..0a8c572d6 --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/03-input-parameters.mdx @@ -0,0 +1,312 @@ +--- +title: Common Input Parameters +--- + +import Citation from '@site/src/components/Citation'; +import CitationFootnote from '@site/src/components/CitationFootnote'; +import NavContainer from '@site/src/components/NavContainer'; +import TableReference from '@site/src/components/TableReference'; +import TableVertical from '@site/src/components/TableVertical'; + + + +# Common Input Parameters + +This chapter documents the inputs that the CLE screening models share. Each model also takes the mechanism-specific flaw inputs documented in its +own manual — compaction categories, settlement and geometry measures, conduit and wall descriptions, climate and surface-cover selections, and the +observation that overrides its flaw probability. The shared inputs fall into seven groups: hydraulic loading, the units and datum convention, the +embankment description, the failure-pathway inputs, the upstream seepage-limiting elements, the initiation material inputs of the compaction models, +and the intervention and breach probabilities. Each group gives the short facts for its quantities in a table, followed by a description of each +input and, for selections, a list of the available choices and what each implies. Where an input applies to only one of the two model groups +(see [The Two Model Groups](./02-model-overview.mdx#the-two-model-groups)), the description says so. + +Several initiation inputs of the compaction models are statistical distributions that characterize uncertainty in a physical quantity. A +distribution input may take any of five forms: + +- **Deterministic** — a single value. +- **Uniform** — a low and a high bound. +- **Normal** — a mean and a standard deviation. +- **Triangular** — a low, a most likely (mode), and a high value. +- **PERT** — a low, a most likely (mode), and a high value. + +Any distribution may also carry a lower limit, an upper limit, or both; a sampled value beyond a limit is set to that limit (the sample is clipped; the distribution is not +renormalized between the limits, and no other treatment of the limits is offered). Every value a distribution can produce +must be positive, so a normal distribution needs a positive lower limit. The Monte Carlo initiation analysis described in +[Initiation](./06-initiation.mdx#delegated-method-cylindrical-pipe-initiation) samples these distributions. + +## Hydraulic Loading + +The hydraulic loading is defined by a headwater series and an optional companion tailwater series +(). + +t'], + ['elevation', 'elevation'], + ['At least two distinct finite values', 'Empty, or one entry per headwater value'], + ['Required', 'No tailwater'], + ]} +/> + +- **Headwater elevation series** — the pool levels at which the event tree is evaluated. The model evaluates every supplied headwater value and + additional pool levels distributed between them (see [Event Tree](./04-event-tree.mdx#headwater-grid)). The series need not be sorted; repeated + values are merged and evaluated once, keeping the tailwater of the first listed pair for that headwater value, and at least two distinct pool + levels must remain. +- **Tailwater elevation series** — the tailwater paired with each headwater value. The series may be left empty when there is no tailwater at any + stage, and any individual entry may be left blank where there is no tailwater at that stage. Tailwater enters only the compaction models, where it + reduces the net head across the flaw (see [Initiation](./06-initiation.mdx#delegated-method-cylindrical-pipe-initiation)); the crest-crack models + report it with the results but do not use it. + +## Units and Datum + +All elevations — headwater, tailwater, crest, flaw, filter, and wall elevations, and the slope-segment elevations — are on one common datum, and +every length is in the same unit as the elevations. Two quantities carry fixed units regardless of the length unit: the critical shear stress in +pounds per square foot and the initial pipe diameter in millimeters. The hydraulic gradient that the initiation analysis forms from the elevations +and the seepage path length is dimensionless, so the initiation result does not depend on the length unit chosen. + +The cylindrical-pipe initiation analysis of the compaction models accepts only positive elevations for the base of the flaw and for the top of the +embankment section it uses (the crest for the shell section, or the core top), and rejects zero or negative values. The datum chosen for a +compaction model must therefore place the embankment above zero elevation. The crest-crack models carry no such restriction. + +:::warning[INPUT NEEDED IN-01 — Applicability] +State the length unit the CLE family is intended to be used in. Several mechanism-specific relationships carry dimensional values with no unit +conversion — for example the tabulated crest-crack depths of the desiccation and frost-cracking models are in feet, and the compaction models band +the flaw base width at fixed numerical thresholds — and only the cross-valley settlement model converts its crack-depth relationship between feet +and meters. The shear stress relationship also uses the unit weight of water in pounds per cubic foot. + +Proposed (unconfirmed): the family is intended for US customary units (elevations and lengths in feet); a user working in SI units should convert +before running any model other than the cross-valley settlement model. Confirm or replace, and state whether this belongs here or in each model's +manual. +::: + +## Embankment Description + +Every model takes the crest elevation and the embankment type (). The compaction models +also describe the embankment cross-section with slope-segment tables, from which they derive the seepage path length through the flaw and, where +a model scores it, the base width and height of the embankment section. + +c', + '—', + 'Wt,s', + 'zb,ks, mks', + 'zt,c, Wt,c', + 'zb,kc, mkc', + ], + ['elevation', 'selection', 'length', 'elevation; horizontal per vertical', 'elevation; length', 'elevation; horizontal per vertical'], + [ + 'Finite; > 0 for the compaction models', + 'One of three types', + '> 0', + 'At least one segment per face', + 'Elevation > 0; width > 0', + 'At least one segment per face', + ], + ['Required', 'Required', 'Required', 'Required', 'Omitted', 'Omitted'], + ]} +/> + +- **Crest elevation** — the top of the embankment. The crest-crack models measure the crack depth down from it; the compaction models use it as + the top of the shell and, where a model applies the normal-pool heuristic, as the top of dam (see + [Flaw Conventions](./05-flaw-conventions.mdx#flaw-elevation)). +- **Embankment type** — the cross-section archetype that sets the crack-filling probability of Node 5 (see + [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#node-5-progression--no-crack-filling)). The choices follow + Fell et al. (2024) : + - **Homogeneous earthfill** — a single earthfill zone. + - **Zoned earthfill** — an earthfill dam with more than one earthfill zone. + - **Central-core earth and rockfill** — an earth core flanked by rockfill shells. +- **Shell top width and face segments** — the outer embankment section. Each face, upstream and downstream, is a list of segments ordered from the + crest downward; the top width is $W_{t,s}$, and segment $k$ is described by the elevation of its bottom $z_{b,k}^{s}$ and its slope $m_k^{s}$, + expressed as horizontal distance per unit of vertical rise. The lowest segment bottom on either face is the base of the section. +- **Core top elevation, top width, and face segments** — an optional inner section, described the same way by its top elevation $z_{t,c}$, its top + width $W_{t,c}$, and its segments $z_{b,k}^{c}$ and $m_k^{c}$. When a core section is supplied it + must be complete, and it replaces the shell for every purpose the section serves — the seepage path length of initiation and any base width and + height the flaw assessment scores — whatever the embankment type. When it is omitted, the shell section is used. The later chapters write + $z_t$ and $W_t$ for the top elevation and top width of the section used by the model: $z_{t,c}$ and $W_{t,c}$ for the core, or the crest + $z_c$ and $W_{t,s}$ for the shell. A low-permeability core flanked + by free-draining shells carries the head loss across the core alone, which is why the core section, when supplied, governs + (see [Initiation](./06-initiation.mdx#seepage-path-length)). + +## Failure-Pathway Inputs + +The failure-pathway inputs describe what the seepage path through the flaw meets on its way downstream: whether a filter intersects it, and which +soils must hold a roof over the enlarging pipe. The two model groups take them in different forms because their flaws differ — a crest crack spans +a range of elevations, while a compaction flaw sits at one elevation +(see [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx)). Each soil is identified by its class in the Unified Soil +Classification System (USCS). + +tf', '—', '—', '—', '—', '—'], + ['elevation', 'USCS classes', 'USCS classes', 'selection', 'USCS classes', 'filter and base soils'], + [ + 'Finite', + 'At least one, unless the filter top reaches the crest', + 'At least one when a filter is present', + 'Present or absent', + 'At least one', + 'Required when a filter is present', + ], + ['No filter', 'Required', 'Omitted', 'Absent', 'Required', 'Omitted'], + ]} +/> + +- **Top-of-filter elevation** (crest-crack models) — the top of the first drain or filter zone the failure pathway meets. Leaving it blank states + that no filter intersects the pathway. When it is supplied, the continuation inputs and the below-filter roof materials are required; when it is + blank, both must be omitted. A filter caps the pathway when its top reaches the crest and no full-pathway roof materials are given: the + below-filter pathway then governs at every pool. When full-pathway materials are supplied, pools above the filter's top are evaluated on the + full pathway even if the filter reaches the crest. +- **Full-pathway roof materials** (crest-crack models) — the soil classifications of every non-filter material along the whole failure pathway, + from the upstream face to the downstream face. They govern roof support whenever the pool is above the filter's top, and at every pool when no + filter is present. The list may be empty only when the filter's top reaches the crest. +- **Below-filter roof materials** (crest-crack models) — the soil classifications of the non-filter materials on the pathway from the upstream face + to the first intersecting filter, taken at the bottom-of-crack elevation. They govern roof support whenever the pool is at or below the filter's + top. + +:::warning[INPUT NEEDED IN-15 — Definition] +The elevation, or range of elevations, at which the full-pathway roof materials are to be taken is not stated. The below-filter materials are taken +at the bottom-of-crack elevation; the full-pathway materials are described only as every non-filter material from the upstream face to the +downstream face. State the elevation or elevation range the user should cut the section at when listing them (for example, the bottom of the +crack, the top of the filter, or the whole depth of the crack above the filter's top). +::: + +- **Filter presence** (compaction models) — whether a filter intersects the failure pathway at the flaw elevation (for example, a diaphragm filter + around a conduit, or a filter zone against a wall or along the foundation contact). The core compaction model expresses this by whether + continuation inputs are supplied. +- **Embankment materials** (compaction models) — the soil classifications of the materials the pipe passes through, which govern roof support at + every pool. +- **Continuation inputs** — the filter description and the base soils upstream of it, evaluated by the shared continuation method documented in + the Continuation (Filter Evaluation) Screening Technical Reference Manual. They are required when a filter is present and used only then. The + crest-crack models reject them when no top-of-filter elevation is given; the conduit, wall or cliff, and foundation or abutment compaction models + ignore them when no filter is stated; and in the core compaction model, supplying them is what states that a filter is present. + +Each soil classification is one of the fifteen USCS classes listed with their roof-support probabilities in +[Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#node-4-progression--roof-support). + +## Upstream Seepage-Limiting Elements + +Every model except the conduit compaction model takes the upstream seepage-limiting elements present and, when one of them is a wall, the +elevation of the wall's top (). The conduit compaction model carries no flow-limitation +node input; its Node 6 is fixed at 1.0 +(see [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#node-6-progression--no-flow-limitation)). + +tw'], + ['selections', 'elevation'], + ['Any number of the six element types', 'Finite; only when a wall is present'], + ['None', 'Wall reaches the crest'], + ]} +/> + +- **Upstream elements present** — any combination of the following, following Fell et al. (2024) . When more + than one is present, the most effective governs: + - **Upstream facing element** — any upstream facing, such as a concrete slab, soil-cement wave protection, a geomembrane, or an asphalt lining. + This is the one element that is not a wall. + - **Sheet pile wall** + - **Modern cutoff wall** + - **Non-cementitious wall** + - **Soil-cement-bentonite wall** + - **Column wall** — jet-grouted or soil-mixed column construction. +- **Top-of-wall elevation** — the top of the cutoff or seepage-barrier wall. Leaving it blank states that the wall reaches the crest, or that no wall + is present. It may be supplied only when a wall is among the elements present. How it gates the walls differs between the two groups (see + [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#node-6-progression--no-flow-limitation)). + +:::warning[INPUT NEEDED IN-02 — Definition] +The wall categories "modern cutoff wall" and "non-cementitious wall" are used without a definition. State what each covers (construction type, +era, or material) so the user can select between them and the soil-cement-bentonite and column-wall categories, or cite the table in Fell et al. +(2024) that defines them. +::: + +## Initiation Material Inputs + +The compaction models take the classification of the flaw material and three optional overrides of the initiation analysis +(). The crest-crack models take none of these inputs. + +c', 'D', 'L'], + ['USCS class', 'lb/ft²', 'mm', 'length'], + ['One of nine classes', 'Positive distribution', 'Positive distribution', '> 0'], + ['Required', 'Range by soil class', 'Range by model', 'Derived from the section'], + ]} +/> + +- **Core soil classification** — the soil in which the flaw would erode. It selects the default critical shear stress range + (see [Initiation](./06-initiation.mdx#critical-shear-stress)). The clean and nearly clean sands and gravels — SP-SM, SW, SP, GP-GM, GW, + and GP, the classes with less than 12 percent fines — are not accepted: the compaction mechanisms are not modeled in these soils. The silty + classes SM, GM, and ML are accepted, although they too can be non-plastic. +- **Critical shear stress** — the hydraulic shear stress at which erosion of the flaw walls begins. Left blank, it defaults to a triangular + distribution over the range for the core soil classification. A site-specific distribution — for example from erodibility testing — replaces the + default. +- **Initial pipe diameter** — the diameter of the idealized cylindrical pipe through the poorly compacted zone. Left blank, it defaults to the + model's triangular range (see [Initiation](./06-initiation.mdx#initial-pipe-diameter)). A site-specific distribution replaces the default. +- **Seepage path length** — the length of the pipe over which the net head is lost. Left blank, it is derived from the embankment section at the + flaw elevation (see [Initiation](./06-initiation.mdx#seepage-path-length)). Supplying a length overrides the derivation; it suits a sloping core, + a pipe that does not run perpendicular to the embankment centerline, or a surveyed path. + +## Intervention and Breach + +int', 'Pbr'], + ['—', '—'], + ['0 to 1', '0 to 1'], + ['1.0', '1.0'], + ]} +/> + +- **Probability of unsuccessful intervention** — the probability that detection and intervention fail to stop the developing erosion (Node 7). + The default of 1.0 takes no credit for intervention. +- **Probability of breach** — the probability that the developing erosion leads to an uncontrolled release of the impounded water (Node 8). The + default of 1.0 treats breach as certain once the earlier nodes are reached. + +Both are single values applied at every pool level. + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/04-event-tree.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/04-event-tree.mdx new file mode 100644 index 000000000..3d93175b0 --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/04-event-tree.mdx @@ -0,0 +1,159 @@ +--- +title: Event Tree +--- + +import Citation from '@site/src/components/Citation'; +import CitationFootnote from '@site/src/components/CitationFootnote'; +import Equation from '@site/src/components/Equation'; +import NavContainer from '@site/src/components/NavContainer'; +import TableReference from '@site/src/components/TableReference'; +import TableVertical from '@site/src/components/TableVertical'; + + + +# Event Tree + +Every CLE screening model computes the system response probability at each headwater level by multiplying the probabilities of the same eight-node +event tree. The tree follows the sequence of the concentrated leak erosion trees in the +[Typical Event Tree Database](/docs/toolbox-technical-manuals/risk-calculations-suite/typical-event-tree-database/v1.5/Concentrated-Leak-Erosion) +— flaw, initiation, continuation, progression, unsuccessful intervention, breach — but no database tree is presented here as the screening tree. +The database trees compare with it as follows: + +- The [homogeneous-embankment](/docs/toolbox-technical-manuals/risk-calculations-suite/typical-event-tree-database/v1.5/Concentrated-Leak-Erosion#homogeneous-embankment) + tree carries progression as a single crack-stability node. +- The [zoned-embankment](/docs/toolbox-technical-manuals/risk-calculations-suite/typical-event-tree-database/v1.5/Concentrated-Leak-Erosion#zoned-embankment) + tree splits the flaw between the core and the shell and adds two progression nodes, asking whether upstream zoning fails to limit flow into the + crack and whether it fails to heal the crack. +- The [soil-rock interface](/docs/toolbox-technical-manuals/risk-calculations-suite/typical-event-tree-database/v1.5/Concentrated-Leak-Erosion#at-soil-rock-interface) + tree, the closest to a foundation or abutment contact flaw, splits the flaw into continuous rock defects and ineffective foundation treatment, + and carries progression as clogging and crack stability, with no crack-filling or flow-limitation node. +- The [embedded-structure](/docs/toolbox-technical-manuals/risk-calculations-suite/typical-event-tree-database/v1.5/Concentrated-Leak-Erosion#along-embedded-structure) + tree, written for a crack along a conduit, carries the same eight nodes as the screening tree, with flow limitation before crack filling. + +The screening tree has one flaw node and always carries three progression nodes — roof support (the crack-stability question), crack filling, +and flow limitation — each evaluated from a table of Fell et al. (2024) . Because the SRP is a product of the +node probabilities, the order of the progression nodes does not affect the result. + +## Node Summary + + summarizes the eight nodes and how each model group populates them. The sections that +follow state the engineering question each node answers and where it is documented in full. + + + +## Pool Dependence + +The nodes that vary with pool level differ between the two groups, because the flaws sit differently in the embankment: + +- **Crest-crack models** — the crack is open from the crest down to its bottom, so the pool can stand anywhere within it. Initiation, continuation, + roof support, and flow limitation are all evaluated at each pool level, because each depends on where the pool stands relative to the bottom of + the crack, the top of the filter, or the top of a wall. Flaw, crack filling, intervention, and breach are the same at every pool level. +- **Compaction models** — the flaw is at one elevation, so the failure pathway, the filter it meets, the materials it passes through, and the + walls that reach it are fixed. Only initiation varies with pool level; the other seven nodes are the same at every pool level. + +## Headwater Grid + +The event tree is evaluated at every supplied headwater value and at additional pool levels distributed between successive supplied values, with +the tailwater at each additional level interpolated linearly between the supplied pairs. An added level in an interval next to a blank tailwater +entry carries no tailwater. The supplied values are sorted, and a repeated headwater +value is evaluated once. The added levels give the system response curve enough resolution to locate the pool levels at which a node steps — the +bottom of a crest crack, the top of a filter, or the top of a wall — between widely spaced supplied values; a pool level of particular interest is +captured exactly by supplying it. + +## Node 1: Flaw + +*Question:* Does a continuous flaw of the model's mechanism — a transverse crack, or a continuous zone of poorly compacted soil — exist through the +embankment? + +The probability $P_{flaw}$ is the result of the model's own flaw assessment, documented in its manual. The common scoring scheme most of the +assessments use, the observation overrides, and the flaw elevation that each passes to the downstream nodes are documented in +[Flaw Conventions](./05-flaw-conventions.mdx). The flaw is a condition of the embankment, so $P_{flaw}$ is the same at every pool level. + +## Node 2: Initiation + +*Question:* Is the hydraulic shear stress on the walls of the flaw large enough to initiate concentrated leak erosion? + +The initiation probability $P_{init}(h)$ is the headwater gate of the crest-crack models or the cylindrical-pipe Monte Carlo analysis of the +compaction models, both documented in [Initiation](./06-initiation.mdx). + +## Nodes 3 to 6: Continuation and Progression + +*Questions:* Does an unfiltered exit allow erosion to continue (Node 3)? Does the material over the enlarging pipe hold a roof (Node 4)? Is the +pipe left unfilled by material from upstream (Node 5)? Does the flow through the pipe remain unrestricted by upstream elements (Node 6)? + +The probabilities $P_{cont}$, $P_{roof}$, $P_{nf}$, and $P_{nl}$ follow from the failure pathway — where it runs, which filter it meets, which +materials it passes through, and which upstream elements reach it — as documented in +[Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx). + +## Nodes 7 and 8: Unsuccessful Intervention and Breach + +*Questions:* Do detection and intervention fail (Node 7)? Does an uncontrolled release of impounded water occur (Node 8)? + +The final two nodes are the user-specified probabilities $P_{int}$ and $P_{br}$ documented in +[Common Input Parameters](./03-input-parameters.mdx#intervention-and-breach). Each is one value applied at every pool level. Both default to 1.0 — +no successful intervention and certain breach — a conservative screening posture. + +## SRP Computation + +The system response probability at each pool level is the product of the eight node probabilities: + + + +where: + +> $P_{flaw}$ = probability of a continuous flaw (Node 1) +> $P_{init}(h)$ = probability of initiation (Node 2) +> $P_{cont}(h)$ = probability of continuation (Node 3) +> $P_{roof}(h)$ = probability of roof support (Node 4) +> $P_{nf}$ = probability of no crack filling (Node 5) +> $P_{nl}(h)$ = probability of no flow limitation (Node 6) +> $P_{int}$ = probability of unsuccessful intervention (Node 7) +> $P_{br}$ = probability of breach (Node 8) + +For the compaction models, $P_{cont}$, $P_{roof}$, and $P_{nl}$ are the same at every pool level. Every model reports an SRP that is never +exactly zero: where the product is zero — for example below the bottom of a crest crack, where initiation is zero — the SRP is reported as a +negligibly small value instead, so that the curve can be carried into calculations that work with the logarithm of probability. + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/05-flaw-conventions.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/05-flaw-conventions.mdx new file mode 100644 index 000000000..f859c088a --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/05-flaw-conventions.mdx @@ -0,0 +1,210 @@ +--- +title: Flaw Conventions +--- + +import Citation from '@site/src/components/Citation'; +import CitationFootnote from '@site/src/components/CitationFootnote'; +import Equation from '@site/src/components/Equation'; +import EquationReference from '@site/src/components/EquationReference'; +import NavContainer from '@site/src/components/NavContainer'; +import TableReference from '@site/src/components/TableReference'; +import TableVertical from '@site/src/components/TableVertical'; + + + +# Flaw Conventions + +Node 1 of every CLE screening model is the probability $P_{flaw}$ that a continuous flaw of the model's mechanism exists through the embankment. +Each model's own manual documents its factors, its category and banding tables, and its flaw-probability curve. This chapter documents the +conventions those assessments share: the risk-factor and likelihood-factor scoring scheme, the optional flaw-continuity stage, the observation +overrides, the treatment of an inapplicable mechanism, and the flaw elevation each assessment passes to the downstream nodes. + +## Risk-Factor and Likelihood-Factor Scoring + +Seven of the eight models — every model except cross-valley differential settlement, which uses a settlement-based relationship documented in its +own manual — estimate the flaw probability by the risk-factor and likelihood-factor scoring of Fell et al. (2024) . +The scheme turns a small number of site factors, each judged on a common scale, into one flaw probability. + +### Factors + +Each model defines its own list of factors. Every factor $j$ carries two numbers: + +- **Risk factor** $RF_j$ — a small integer weight expressing how strongly the factor influences the likelihood of a flaw. Six of the seven + scoring models weight three factors 3, 2, and 1; the conduit compaction model weights four factors 3, 2, 2, and 2. +- **Likelihood factor** $LF_j$ — the factor's score for the site, on a scale on which 1 is the least likely condition for a flaw and 4 the most + likely. A factor's likelihood factor comes either from a category the user selects (for example a compaction category or a climate) or from a + banding rule that places a measured quantity (for example a width-to-height ratio or a settlement ratio) into one of the bands. Some categories + — typically Unknown, or a condition between two bands — carry intermediate scores such as 2.5 or 3.5. + +### Weighted sum + +The scores combine into the weighted sum + + + +which rises as more of the factors, and the more heavily weighted factors, point toward a flaw. + +### Flaw probability from the weighted sum + +The weighted sum is converted to a flaw probability by a curve tabulated at a few points $(\Sigma_k, P_k)$ for each model. Between two tabulated +points the logarithm of the probability is interpolated linearly in the weighted sum: + + + +so that equal steps in the weighted sum multiply the probability by equal factors. A weighted sum beyond either end of the tabulated range takes +the probability at that end. In several models the lowest tabulated point carries a negligibly small probability, so that the probability falls +steeply toward the low end of the tabulated range. + +Two further rules can force the raw flaw probability $P_{raw}$ to zero before the curve is consulted: + +- **Incompatible-condition factors** — a factor may be marked as one whose least likely score (a likelihood factor of exactly 1) describes a + condition physically incompatible with a continuous flaw. When any such factor scores 1, the flaw probability is zero whatever the other factors + score. The wall or cliff settlement model marks two of its geometric factors this way, and the foundation or abutment compaction model marks its + foundation-preparation factor. +- **Lower threshold** — a model may set a weighted sum below which the flaw probability is zero. The desiccation and frost-cracking models use a + threshold of 13: a weighted sum below 13 gives a zero flaw probability, while a sum of exactly 13 takes the first tabulated value. + +A flaw probability forced to zero by either rule leaves the SRP at the negligible value described in +[Event Tree](./04-event-tree.mdx#srp-computation), unless an observation override applies: in the desiccation and frost-cracking models a reported +observation sets the flaw probability to 1.0 even when the weighted sum is below the threshold of 13 (see +[Observation Overrides](#observation-overrides)). Only an inapplicable mechanism sets the override aside +(see [Inapplicable Mechanism](#inapplicable-mechanism)). + +:::warning[INPUT NEEDED IN-03 — Engineering Basis] +The conversion of the weighted sum to a flaw probability by interpolating the logarithm of the probability linearly between tabulated points +(), holding the end value beyond the tabulated range, and — in several models — +anchoring the lowest tabulated point at a negligibly small probability, requires an engineering basis. State whether Fell et al. (2024) presents +the weighted-sum-to-probability relationships as curves to be interpolated in this way (and cite the figure or table), or provide the +engineering-judgment attribution. + +Proposed (unconfirmed): the tabulated probabilities span several orders of magnitude, so interpolation in the logarithm of probability is the +natural reading of a relationship plotted on a logarithmic probability axis; the negligible anchor reproduces a relationship that falls toward +zero at the lowest attainable score. Confirm or replace. +::: + +## Flaw-Continuity Stage + +Three compaction models — poor compaction within the core, adjacent to a wall or cliff, and along the foundation or abutment — apply a second +stage after the scoring. The raw flaw probability is multiplied by a flaw-continuity probability $P_{cf}$, interpolated in the same log-linear +manner from the likelihood factor of the model's compaction factor: + + + +where $LF_{comp}$ is the likelihood factor of the core-compaction category (core compaction), of the wall-fill compaction category (wall or cliff +compaction), or of the contact-zone compaction method (foundation or abutment compaction). The three models share the flaw-continuity curve of +. + +-4', '1 × 10-3', '1 × 10-1'], + ]} +/> + +A fractional compaction score, such as the 3.5 assigned to an Unknown core-compaction category, is interpolated between the tabulated points. The +other four scoring models (wall or cliff settlement, conduit compaction, desiccation, and frost cracking) use the scored probability directly: +$P_{flaw} = P_{raw}(\Sigma)$. + +:::warning[INPUT NEEDED IN-04 — Engineering Basis] +The flaw-continuity stage and its curve (: likelihood factor 1, 2, 3, 4 → +negligible, 1 × 10-4, 1 × 10-3, 1 × 10-1) require an engineering basis and a definition. State what the +flaw-continuity probability represents physically (for example, the probability that a poorly compacted zone is continuous across the full +width of the embankment, given that it exists) and cite its source, or provide the engineering-judgment attribution. Confirm also that one curve +is intended for all three models, keyed to a different compaction factor in each. +::: + +## Observation Overrides + +Most models accept an observation that demonstrates the flaw directly. When the observation is reported, the flaw probability is set to 1.0 whatever +the factors score; in the three compaction models with an override, the factor selections also become optional. The observations are: + +- **Cross-valley differential settlement** — continuous transverse cracking observed. +- **Desiccation cracking and frost cracking** — transverse cracking observed and judged continuous. +- **Poor compaction within the core** — embankment leakage or wet spots observed. The observed elevation of the leakage or wet spots also becomes + the flaw elevation (see [Flaw Elevation](#flaw-elevation)). +- **Poor compaction adjacent to a conduit** — leakage observed adjacent to the conduit. The flaw elevation is unchanged. +- **Poor compaction adjacent to a wall or cliff** — leakage observed adjacent to the wall. The flaw elevation is unchanged. + +The wall or cliff settlement and the foundation or abutment compaction models have no observation override. In the foundation or abutment +compaction model, a well-prepared foundation (a foundation-preparation likelihood factor of 1) instead forces the flaw probability to zero, as +described under [incompatible-condition factors](#flaw-probability-from-the-weighted-sum). + +The override sets only Node 1. Every downstream node is evaluated as usual, so an observed flaw still needs the pool to reach it, an unfiltered +exit, and the progression conditions before it contributes to the SRP. + +## Inapplicable Mechanism + +The desiccation and frost-cracking models score the soils the crest crack would pass through. When every listed crest soil is one in which the +mechanism cannot occur (one of the clean or nearly clean sands and gravels SP-SM, SW, SP, GP-GM, GW, and GP, the classes with less than +12 percent fines), the mechanism does not apply: no crack forms, the flaw probability is negligible, the observation +override is not consulted, and initiation is zero at every pool level (see [Initiation](./06-initiation.mdx#crest-crack-models-headwater-gate)). +The soil sets are documented in those models' manuals. + +## Flaw Elevation + +Each flaw assessment passes one elevation to the downstream nodes along with the flaw probability. It fixes where erosion can begin, where the +failure pathway runs, and which filter and upstream walls the pathway meets. + +### Crest-crack models: bottom of the crack + +The flaw of a crest-crack model is a transverse crack open at the crest. The flaw assessment estimates the crack depth $d_c$ — from the settlement +and geometry of the mechanism, or from tabulated depths by climate and soil — and the bottom of the crack is + + + +where $z_c$ is the crest elevation. Water can enter the crack only when the pool rises above $z_{bc}$, which is the initiation gate of +[Initiation](./06-initiation.mdx#crest-crack-models-headwater-gate). When an inapplicable mechanism produces no crack, the crack depth is zero and +$z_{bc}$ equals the crest elevation. + +### Compaction models: base of the flaw + +The flaw of a compaction model is idealized as a cylindrical pipe whose invert is the base-of-flaw elevation $z_f$. The four models resolve it in +three ways: + +- **Normal-pool heuristic** — for poor compaction within the core, and for poor compaction along an abutment contact, the base of the flaw is + placed one third of the way from the normal pool elevation $z_{np}$ to the crest: + + + + The crest must be above the normal pool. For the core model, an observation of leakage or wet spots replaces this with the observed elevation. +- **Direct input** — for poor compaction adjacent to a conduit, the base of the flaw is the bottom-of-conduit elevation; for poor compaction along + a foundation contact, it is the embankment–foundation contact elevation. +- **Derived from the wall height** — for poor compaction adjacent to a wall or cliff, the base of the flaw is the crest elevation less the wall or + cliff height $h_w$, $z_f = z_c - h_w$. The model also requires it to lie at or above the lowest segment bottom of the shell section. + +In all four models, the base of the flaw may not lie above the top of the embankment section used for initiation (the core section when one is +supplied, otherwise the shell), nor below the lowest segment bottom of either face of that section, and it must be a positive elevation (see +[Units and Datum](./03-input-parameters.mdx#units-and-datum)). + +:::warning[INPUT NEEDED IN-05 — Engineering Basis] +The normal-pool heuristic of — placing the base of a core or abutment compaction +flaw one third of the way from the normal pool to the crest — requires an engineering basis. Provide the literature source or the +engineering-judgment attribution to state here, including why the normal pool and not the base of the embankment is the lower reference. + +Proposed (unconfirmed): a screening convention that places the flaw within the band of pools the embankment experiences, so that initiation is +evaluated above normal pool where loading is frequent. Confirm or replace. +::: + +### Section geometry used in scoring + +Where a compaction model scores the geometry of the embankment section — the core compaction and the foundation or abutment compaction models band +the base width $W$ and the width-to-height ratio $W/H_s$ — the two dimensions come from the section used by the model (the core section when +supplied, otherwise the shell; see [Common Input Parameters](./03-input-parameters.mdx#embankment-description)): + + + +where $H_s$ is the height of the section; $z_t$ and $W_t$ are its top elevation and top width — the core top $z_{t,c}$ and core top width +$W_{t,c}$ for the core section, or the crest $z_c$ and shell top width $W_{t,s}$ for the shell; $z_{base}$ is the lowest segment bottom on +either face; and $S_{us}$ and $S_{ds}$ are the horizontal spans of the upstream and downstream faces from the top down to $z_{base}$, each the sum over its segments of the +vertical extent times the segment slope. Both models accept a user-supplied base width and height in place of the derived values; the supplied +values replace them in the scoring only, while the initiation analysis keeps the section geometry. + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/06-initiation.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/06-initiation.mdx new file mode 100644 index 000000000..0f93595ec --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/06-initiation.mdx @@ -0,0 +1,234 @@ +--- +title: Initiation +--- + +import Equation from '@site/src/components/Equation'; +import EquationReference from '@site/src/components/EquationReference'; +import NavContainer from '@site/src/components/NavContainer'; +import TableReference from '@site/src/components/TableReference'; +import TableVertical from '@site/src/components/TableVertical'; + + + +# Initiation + +Node 2 of every CLE screening model asks whether the hydraulic shear stress exerted on the walls of the flaw by the water flowing through it is +large enough to begin eroding them. The two model groups answer it differently. The crest-crack models apply a headwater gate at the bottom of the +open crack. The compaction models hand the question to the cylindrical-pipe initiation analysis of the Concentrated Leak Erosion Initiation +toolbox, evaluated by Monte Carlo simulation over the soil's critical shear stress and the initial pipe diameter. + +## Crest-Crack Models: Headwater Gate + +**Models:** cross-valley differential settlement, settlement adjacent to a wall or cliff, desiccation cracking, and frost cracking. + +The flaw is a transverse crack open from the crest down to the bottom-of-crack elevation $z_{bc}$ +(see [Flaw Conventions](./05-flaw-conventions.mdx#crest-crack-models-bottom-of-the-crack)). Water enters the crack only once the pool rises above +its bottom, and once it does, initiation is taken as certain: + + + +The gate turns on strictly above the bottom of the crack: at a pool exactly at $z_{bc}$, initiation is zero. No shear stress is computed, no +cylindrical-pipe analysis is performed, and tailwater plays no part. In the desiccation and frost-cracking models, an inapplicable mechanism (see +[Flaw Conventions](./05-flaw-conventions.mdx#inapplicable-mechanism)) holds initiation at zero at every pool level. + +The gate makes the crack depth the controlling quantity of the crest-crack models' initiation: the deeper the crack, the lower the pool at which +the SRP steps from negligible to its full value. Above $z_{bc}$, the SRP of a crest-crack model varies with pool level only through the +continuation, roof-support, and flow-limitation nodes (see [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx)). + +:::warning[INPUT NEEDED IN-06 — Engineering Basis] +The initiation probability of 1.0 applied once the pool rises above the bottom of an open crest crack +() requires an engineering basis to state here. The model records two different +attributions for it: one describes it as a conservative screening-level convention that is not an explicit finding of Pierson (2023), while +others attribute the certain initiation of a transverse crest crack to Pierson (2023). Resolve which is intended. If Pierson (2023) is to be +cited, provide its full bibliographic details (authors, title, publisher or journal, and report number) so that it can be added to the +bibliography. + +Proposed (unconfirmed): an open transverse crack at the crest is wide enough that the shear stress of any flow through it exceeds the critical +shear stress of the crack walls, so initiation is taken as certain once water can enter the crack — a conservative screening posture under which a +site needing a lower value would adjust the SRP. Confirm or replace. +::: + +## Delegated Method: Cylindrical Pipe Initiation + +**Node served:** Node 2 (Initiation) of the compaction models — poor compaction within the core, adjacent to a conduit, adjacent to a wall or +cliff, and along the foundation or abutment. + +**Question answered:** At each pool level, does the flow through a pipe of initial diameter $D$ at the base of the flaw exert a hydraulic shear +stress $\tau$ on the pipe wall that exceeds the critical shear stress $\tau_c$ of the soil? A poorly compacted zone is idealized as such a pipe, +and the cylindrical-pipe analysis compares the two stresses over the uncertainty in $\tau_c$ and $D$. The theory is in the +[Cylindrical Pipe](/docs/toolbox-technical-manuals/internal-erosion-suite/concentrated-leak-erosion-initiation/v1.0/cylindrical-pipe) chapter +of the Concentrated Leak Erosion Initiation manual; this section documents what the screening models supply to it, what they take from it, and +each screening-level choice. + +### Governing relationships + +The pipe invert is the base-of-flaw elevation $z_f$ resolved by the flaw assessment +(see [Flaw Conventions](./05-flaw-conventions.mdx#compaction-models-base-of-the-flaw)). At each pool level $h$ with tailwater $h_t$, the upstream +and downstream heads above the invert are each limited to be no less than zero: + + + +where $L$ is the seepage path length and $i$ the average hydraulic gradient along the pipe. A pool at or below the invert gives no flow and no +initiation; a tailwater below the invert, or no tailwater, gives no submergence. The hydraulic shear stress on the wall of a cylindrical pipe +carrying this gradient is, as derived in the toolbox manual's +[Hydraulic Shear Stress](/docs/toolbox-technical-manuals/internal-erosion-suite/concentrated-leak-erosion-initiation/v1.0/cylindrical-pipe#hydraulic-shear-stress) +section, + + + +with $\gamma_w$ the unit weight of water (62.4 lb/ft³) and $D$ converted from millimeters to feet. Erosion initiates in a realization when +$\tau > \tau_c$. The initiation probability at each pool level is the fraction of Monte Carlo realizations in which it does, as in the toolbox +manual's +[Likelihood of Initiation](/docs/toolbox-technical-manuals/internal-erosion-suite/concentrated-leak-erosion-initiation/v1.0/cylindrical-pipe#likelihood-of-initiation-of-concentrated-leak-erosion) +section: + + + +where each of the $N$ realizations draws one critical shear stress $\tau_{c,n}$ and one initial pipe diameter $D_n$ and holds them at every pool +level. Because the same draws are carried across pool levels, the initiation probability does not decrease wherever the gradient rises with pool +level: it rises, or stays flat at 0, at 1, or at a plateau between them. It can fall only where a rising tailwater reduces the net head across the +pipe faster than the pool raises it. + +### Seepage path length + +Unless the user supplies a seepage path length, the model measures the pipe as a horizontal line at the invert elevation across the embankment +section — the core section when one is supplied, the shell section otherwise (see +[Common Input Parameters](./03-input-parameters.mdx#embankment-description)): + + + +where $W_t$ is the top width of the section ($W_{t,c}$ for the core, $W_{t,s}$ for the shell) and $S_{us}(z_f)$ and $S_{ds}(z_f)$ are the +horizontal spans of the upstream and downstream faces from the top of the section down to the invert. Each span is the sum, over the face +segments, of the vertical extent of each segment above the invert (only the part above it, for the segment the invert cuts) times the segment's +slope. This is the measurement the toolbox manual prescribes in its +[Core Geometry](/docs/toolbox-technical-manuals/internal-erosion-suite/concentrated-leak-erosion-initiation/v1.0/cylindrical-pipe#core-geometry) +and Hydraulic Shear Stress sections: across the embankment slopes for a homogeneous embankment, and across the impervious core for a zoned +embankment whose outer zones are very permeable. As the manual recommends, a user-supplied length replaces the derivation for a sloping core or a +pipe that does not run perpendicular to the embankment centerline; a surveyed path length may be supplied the same way. The length is the same at every pool level. + +### Critical shear stress + +Left blank, the critical shear stress is a triangular distribution whose low, most likely, and high values are taken from + for the core soil classification. All four compaction models share this table. The +toolbox manual's +[Critical Shear Stress and Initial Pipe Diameter](/docs/toolbox-technical-manuals/internal-erosion-suite/concentrated-leak-erosion-initiation/v1.0/cylindrical-pipe#critical-shear-stress-and-initial-pipe-diameter) +section directs the estimation of a site-specific critical shear stress to the +[Erodibility Parameters](/docs/toolbox-technical-manuals/internal-erosion-suite/erodibility-parameters/v1.0/preface) toolbox; a site-specific +distribution supplied by the user replaces the default. + + + +The six clean and nearly clean sand and gravel classes — SP-SM, SW, SP, GP-GM, GW, and GP, the classes with less than 12 percent fines — have +no entry: the compaction models do not accept them as the core soil classification. The silty classes SM, GM, and ML, which can also be +non-plastic, are accepted. + +:::warning[INPUT NEEDED IN-07 — Engineering Basis] +The default critical shear stress ranges of require a literature source or an +engineering-judgment attribution. Provide the full bibliographic details of the source table (for example, the erodibility data set or +correlation the ranges are taken from) so that it can be added to the bibliography. + +Proposed (unconfirmed): the tabulated values are conversions of round values in pascals — 1, 2, 4, 5, 10, 20, 25, and 50 Pa — which suggests +a source table published in SI units. Confirm or replace. +::: + +:::warning[INPUT NEEDED IN-08 — Decision] +Two entries of do not follow the pascal conversion of the rest of the table: the low +value for GM is 0.021 lb/ft² where 1 Pa is 0.02089 lb/ft², and the high value for GC is 0.418 lb/ft² where 20 Pa is 0.4177 lb/ft². State whether +these two values are intended as printed or should match the converted values used elsewhere in the table. The difference does not change any +result materially but determines what this table should state. +::: + +:::warning[INPUT NEEDED IN-09 — Engineering Basis] +The exclusion of the six clean and nearly clean sand and gravel classes (SP-SM, SW, SP, GP-GM, GW, GP) from all four compaction models requires an engineering basis to +state here — why a poorly compacted zone in these soils is not modeled as a concentrated leak erosion flaw. + +Proposed (unconfirmed): a cohesionless soil cannot sustain an open pipe or crack, so a compaction flaw in it would collapse rather than carry a +concentrated leak; such soils are addressed by other internal-erosion failure modes. Confirm or replace. +::: + +:::warning[INPUT NEEDED IN-10 — Engineering Basis] +The choice of a triangular distribution over the tabulated low, most likely, and high values as the default critical shear stress distribution +requires an engineering basis or an engineering-judgment attribution. + +Proposed (unconfirmed): the triangular form follows the probabilistic input of the Concentrated Leak Erosion Initiation toolbox, which represents +the critical shear stress and the initial pipe diameter by triangular distributions over their minimum, most likely, and maximum values. Confirm +or replace. +::: + +### Initial pipe diameter + +Left blank, the initial pipe diameter is a triangular distribution set by the model (). +The toolbox manual's +[Critical Shear Stress and Initial Pipe Diameter](/docs/toolbox-technical-manuals/internal-erosion-suite/concentrated-leak-erosion-initiation/v1.0/cylindrical-pipe#critical-shear-stress-and-initial-pipe-diameter) +section directs the estimation of an initial opening to the +Concentrated Leak Erosion (Cracking) +toolbox and, because the initial diameter is very difficult to predict, recommends sensitivity analysis; a site-specific distribution supplied by +the user replaces the default. + + + +:::warning[INPUT NEEDED IN-11 — Engineering Basis] +The default initial pipe diameter ranges of require an engineering basis, including +why the conduit model's range is larger than the core model's, and why the wall or cliff model's range has its most likely value at its low +value. Provide a literature source or the engineering-judgment attribution to state here. +::: + +### Screening-level adjustments + +Relative to the toolbox method, the compaction models: + +1. **Use the cylindrical-pipe geometry only.** The horizontal, vertical rectangular, and vertical triangular crack geometries of the toolbox are + not offered; a poorly compacted zone is idealized as a pipe. +2. **Always run the probabilistic analysis.** The toolbox's deterministic option, which compares the most likely values only, is not used, so the + initiation probability always reflects the uncertainty in the critical shear stress and the initial pipe diameter. +3. **Supply default distributions** for the critical shear stress and the initial pipe diameter, which the user may replace. +4. **Derive the seepage path length** from the screening section at the flaw elevation, which the user may replace. + +The toolbox's further results — the critical pipe diameter, the headwater level for initiation, and the steady flow through the pipe — do not +enter the event tree. diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/07-failure-pathway-and-progression.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/07-failure-pathway-and-progression.mdx new file mode 100644 index 000000000..76cd04e9b --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/07-failure-pathway-and-progression.mdx @@ -0,0 +1,205 @@ +--- +title: Failure Pathway and Progression +--- + +import Citation from '@site/src/components/Citation'; +import CitationFootnote from '@site/src/components/CitationFootnote'; +import Equation from '@site/src/components/Equation'; +import NavContainer from '@site/src/components/NavContainer'; +import TableReference from '@site/src/components/TableReference'; +import TableVertical from '@site/src/components/TableVertical'; + + + +# Failure Pathway and Progression + +Nodes 3 through 6 of the CLE event tree depend on what the eroding flaw meets on its way through the embankment: whether a filter intercepts it, +which soils must hold a roof over the enlarging pipe, whether material from upstream can fill it, and whether an upstream element restricts the +flow into it. This chapter documents the failure-pathway conventions that decide those questions and how each of the four nodes is evaluated. + +## The Failure Pathway + +The failure pathway is the route the concentrated leak takes through the embankment, from the upstream face to the downstream face. Its form follows +from the flaw: + +- **Crest-crack models** — the pathway runs along the transverse crack, which is open over a range of elevations from its bottom $z_{bc}$ up to the + crest. The first drain or filter zone the pathway meets defines the top-of-filter elevation $z_{tf}$. While the pool is at or below $z_{tf}$, the + water entering the crack reaches the filter: this is the **below-filter pathway**, from the upstream face to the filter, taken at the + bottom-of-crack elevation. Once the pool rises above $z_{tf}$, the water entering the upper part of the crack passes over the filter's top: this + is the **full pathway**, from the upstream face to the downstream face. A filter **caps the pathway** when its top reaches the crest and no + full-pathway materials are given: the below-filter pathway then governs at every pool level. If full-pathway materials are supplied, pools + above the filter's top are evaluated on the full pathway. With no filter on the pathway, the full pathway governs at every pool level. +- **Compaction models** — the pathway runs through the embankment section at the base-of-flaw elevation $z_f$. The flaw, and so the pathway, is at + one elevation: whether a filter intercepts it, which materials it passes through, and which walls reach it are the same at every pool level. + +The user resolves the pathway against the embankment section and supplies the result — the top-of-filter elevation, the filter and base-soil +description, and the material lists — as documented in [Common Input Parameters](./03-input-parameters.mdx#failure-pathway-inputs). + +## Node 3: Continuation + +*Question:* Does an unfiltered seepage exit exist that allows erosion to continue? + +Where a filter intercepts the pathway, the probability that erosion continues past it, $P_{ce}$, is evaluated by the shared continuation method +from the filter and base-soil description. The method, its alternative evaluation routes, and its basis are documented in the Continuation (Filter +Evaluation) Screening Technical Reference Manual; it draws on the toolbox methods of the +[Filter Evaluation (Continuation)](/docs/toolbox-technical-manuals/internal-erosion-suite/filter-evaluation-continuation/v1.0/general-overview) +manual. Where no filter intercepts the pathway, the exit is unfiltered and continuation is certain. + +**Crest-crack models.** Continuation follows the pathway that governs at each pool level: + + + +Above the top of a filter that does not cap the pathway, the water in the upper part of the crack passes over the filter and exits unfiltered, so +continuation is certain there. A crest-crack SRP curve with a filter therefore steps where the pool crosses $z_{tf}$: continuation rises from +$P_{ce}$ to 1.0, but roof support switches at the same level from the below-filter minimum to the full-pathway minimum, which can be much lower +(for example 0.01 for a downstream rockfill shell classified as GP on the full pathway). The SRP steps up where $P_{ce}$ is smaller than the ratio of the +full-pathway to the below-filter roof support, and steps down where it is larger (see +[Node 4: Progression — Roof Support](#node-4-progression--roof-support)). + +**Compaction models.** Continuation is one value for every pool level: $P_{ce}$ when a filter intercepts the pathway at the flaw elevation, and +1.0 when none does. The continuation method is not evaluated when no filter is present. + +:::warning[INPUT NEEDED IN-12 — Engineering Basis] +The crest-crack pathway convention requires an engineering basis to state here: (1) continuation and roof support share one gate at the top of the +first filter the pathway meets, so that above that elevation erosion is taken to exit unfiltered over the filter's top with certain continuation; +(2) the below-filter roof materials are taken at the bottom-of-crack elevation; and (3) a filter whose top reaches the crest, with no full-pathway +materials given, caps the pathway at every pool level. Provide the literature source or the engineering-judgment attribution. + +Proposed (unconfirmed): a crest crack is open over its full depth, so a pool above the filter's top drives flow through the upper part of the +crack over the filter, where nothing filters the exit; below the filter's top, all flow through the crack reaches the filter. Confirm or replace. +::: + +## Node 4: Progression — Roof Support + +*Question:* Does a continuous stable roof form over the enlarging pipe, so that it stays open? + +Each soil classification on the pathway maps to a roof-support probability per , +following Fell et al. (2024) . Every material along the pathway must hold a roof for the pipe to progress, so +the least capable material governs and the node takes the minimum across the listed materials: + + + +where $\mathcal{M}$ is the material set that applies: + +- **Crest-crack models** — the below-filter materials at pool levels at or below $z_{tf}$ (and at every pool level when the filter caps the + pathway), and the full-pathway materials above it (and at every pool level when no filter is present). Roof support can therefore change where + the pool crosses $z_{tf}$, in step with continuation. +- **Compaction models** — the embankment materials the pipe passes through, at every pool level. + + + +## Node 5: Progression — No Crack Filling + +*Question:* Does the enlarging pipe remain unfilled, rather than being filled and healed by material carried in from upstream? + +The probability $P_{nf}$ depends only on the embankment type, per , following Fell +et al. (2024) . It is the same at every pool level for both groups. + + + +:::warning[INPUT NEEDED IN-13 — Definition] +Confirm the physical meaning of the values in . The model's source table is described +as giving the probability of crack filling, while the event-tree node multiplies the tabulated value directly as the probability of no crack +filling. The two readings agree only for the central-core earth and rockfill value of 0.5; for the earthfill types, a tabulated 1.0 read as the +probability of crack filling would mean certain healing. State which reading the source table supports, cite its table number in Fell et al. +(2024). + +Proposed (unconfirmed): the tabulated value is the probability that crack filling does not stop the erosion — an earthfill embankment has no +upstream zone of material able to wash into and fill the pipe, so no credit is taken, while the upstream zone of a central-core earth and +rockfill embankment is credited with a one-in-two chance of filling it. Confirm or replace. +::: + +## Node 6: Progression — No Flow Limitation + +*Question:* Does the flow into the enlarging pipe remain unrestricted by upstream seepage-limiting elements? + +Each upstream element maps to a probability of no flow limitation per , following Fell +et al. (2024) . When more than one element applies, the most effective one controls the flow, so the node takes +the minimum across the applicable elements; with no applicable element, flow is not limited and $P_{nl} = 1.0$: + + + + + +### Walls that stop below the crest + +A wall limits flow only over its own height. When the user supplies the top-of-wall elevation $z_{tw}$, the walls apply only where the leak +passes beneath their top, while the upstream facing element, which is not a wall, applies wherever it is present. The applicable set $\mathcal{E}$ +is therefore: + +- **Crest-crack models** — gated per pool level. At pool levels at or below $z_{tw}$ every listed element applies; above it the walls drop out + and only the upstream facing element remains, because the flow entering the crack above the wall's top passes over it. +- **Compaction models** — gated once, against the flaw elevation. When $z_{tw} \ge z_f$ the wall reaches the flaw and every listed element applies; + when $z_{tw} < z_f$ the walls drop out and only the upstream facing element remains. Either way, one value applies at every pool level. + +With no top-of-wall elevation, the walls are taken to reach the crest and every listed element applies at every pool level. + +The conduit compaction model carries no flow-limitation input: it models no upstream element, and its Node 6 is fixed at 1.0. + +:::warning[INPUT NEEDED IN-14 — Engineering Basis] +The flow-limitation conventions require an engineering basis to state here: (1) the minimum across several upstream elements governs; (2) a wall +stops limiting flow above its top — per pool level for a crest crack, and against the base of the flaw for a compaction flaw — while an upstream +facing element applies at every pool level; and (3) the conduit compaction model takes no flow-limitation credit at all. Provide the literature +source (including the table in Fell et al. (2024) that gives the values of +) or the engineering-judgment attribution. +::: + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/08-assumptions-and-limitations.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/08-assumptions-and-limitations.mdx new file mode 100644 index 000000000..8e8907824 --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/08-assumptions-and-limitations.mdx @@ -0,0 +1,159 @@ +--- +title: Assumptions and Limitations +--- + +import Citation from '@site/src/components/Citation'; +import CitationFootnote from '@site/src/components/CitationFootnote'; +import NavContainer from '@site/src/components/NavContainer'; + + + +# Assumptions and Limitations + +The CLE screening models apply deliberate simplifications relative to a detailed internal-erosion risk analysis. They reduce input requirements +and are designed to avoid false negatives at the screening level. This chapter lists the assumptions the eight models share; each model's own +manual lists the assumptions of its flaw assessment. Each assumption below states its engineering basis, its effect on results, and the conditions +under which it may not hold. Where an engineering basis is still awaiting confirmation, the assumption cross-references the input marker at its +primary location in this document. + +## Event Tree + +1. **The SRP is the product of the eight node probabilities.** Each node is the probability of its event given that every earlier node has + occurred (see [Event Tree](./04-event-tree.mdx#srp-computation)). + - *Engineering basis:* Event-tree convention for internal-erosion failure modes, following the node sequence of the concentrated leak erosion + trees of the Typical Event Tree Database and the progression conditions of Fell et al. (2024) . + - *Effect on results:* The SRP can be no larger than its smallest node, so a single low node — a filter, a cohesionless roof material, an + effective cutoff — can control the curve. + - *When it may not hold:* Where the nodes are strongly dependent in ways the conditional values do not capture, for example where the same soil + property controls both initiation and roof support; a detailed analysis then assesses the joint condition directly. +2. **Intervention and breach are single values that apply at every pool level, defaulting to 1.0.** + - *Engineering basis:* Screening design decision (engineering judgment): taking no credit for intervention and treating breach as certain is the + conservative screening posture, and site-specific values can be supplied. + - *Effect on results:* With the defaults, the SRP is the probability of reaching the end of the progression nodes. + - *When it may not hold:* Where monitoring and response time, or breach resistance, vary strongly with pool level; a pool-dependent assessment + belongs in a detailed analysis. +3. **The curve is resolved on a grid of pool levels.** The tree is evaluated at every supplied headwater value and at intermediate levels between + them (see [Event Tree](./04-event-tree.mdx#headwater-grid)). + - *Engineering basis:* Screening simplification (engineering judgment): the nodes that step with pool level — at the bottom of a crest crack, the + top of a filter, or the top of a wall — are located to within one grid interval. + - *Effect on results:* A step between two grid levels appears in the curve as a change across that interval rather than at its exact elevation. + - *When it may not hold:* Where the exact pool level of a step matters; supplying that pool level as a headwater value places a grid level on it. + +## Flaw + +4. **The flaw probability is the same at every pool level.** + - *Engineering basis:* Engineering judgment: the existence of a flaw is a condition of the embankment, not of the loading. + - *Effect on results:* Pool dependence of the SRP comes entirely from initiation and, for the crest-crack models, from the pathway gates. + - *When it may not hold:* Where cracking itself depends on pool level, such as hydraulic fracture under rising pool; that mechanism is outside + the family. +5. **Qualitative site factors are converted to a flaw probability by risk-factor and likelihood-factor scoring and a log-linear curve.** This applies + to every model except cross-valley differential settlement (see [Flaw Conventions](./05-flaw-conventions.mdx#risk-factor-and-likelihood-factor-scoring)). + - *Engineering basis:* Fell et al. (2024) for the scoring scheme; the interpolation form is requested at + marker IN-03 in [Flaw Conventions](./05-flaw-conventions.mdx#flaw-probability-from-the-weighted-sum). + - *Effect on results:* The flaw probability moves by roughly constant factors per step of the weighted sum, and takes the end value beyond the + tabulated range. + - *When it may not hold:* Where site investigation supports a flaw probability more directly than the factor scores do; the risk team's + estimate then replaces the screening value. +6. **Three compaction models multiply the scored probability by a flaw-continuity probability keyed to their compaction factor.** + - *Engineering basis:* Basis requested — see marker IN-04 in [Flaw Conventions](./05-flaw-conventions.mdx#flaw-continuity-stage). + - *Effect on results:* The best compaction rating (a likelihood factor of 1) reduces the flaw probability to negligible, and each step toward + poorer compaction above a likelihood factor of 2 raises it by one to two orders of magnitude. The compaction category enters the flaw + probability twice — through its weight in the weighted sum and again through the flaw-continuity probability — so the flaw probability is + especially sensitive to that selection. + - *When it may not hold:* Where compaction records conflict with field performance, such as observed seepage through an embankment rated well + compacted; the observation override then applies. +7. **An observed flaw has a flaw probability of 1.0.** + - *Engineering basis:* Engineering judgment: an observation of leakage, wet spots, or continuous cracking demonstrates the flaw, so its + probability is not estimated from site factors. + - *Effect on results:* The SRP is governed entirely by the downstream nodes. + - *When it may not hold:* Where the observation has another explanation, such as surface cracking that is not continuous or seepage along a + different path; the override should be reserved for observations judged to demonstrate the modeled flaw. +8. **The base of a core or abutment compaction flaw is placed one third of the way from the normal pool to the crest**, unless an observed leakage + elevation (core) replaces it. + - *Engineering basis:* Basis requested — see marker IN-05 in [Flaw Conventions](./05-flaw-conventions.mdx#compaction-models-base-of-the-flaw). + - *Effect on results:* Initiation cannot occur, and the SRP is negligible, at pool levels at or below this elevation, and walls are judged + against it. Filter presence is stated by the user, not compared with this elevation. + - *When it may not hold:* Where construction records or performance locate the poorly compacted zone; an observed leakage elevation should then + be used where the model accepts one. + +## Initiation + +9. **Initiation in a crest crack is certain once the pool rises above the bottom of the crack, and zero at or below it.** Tailwater plays no part. + - *Engineering basis:* Basis requested — see marker IN-06 in [Initiation](./06-initiation.mdx#crest-crack-models-headwater-gate). + - *Effect on results:* The crack depth alone sets the pool level at which the crest-crack SRP rises from negligible; no shear stress analysis can + lower the initiation probability above that level (conservative). + - *When it may not hold:* For narrow cracks in highly erosion-resistant soils, or cracks that close or fill below the surface; a crack-geometry + initiation analysis then applies. +10. **A poorly compacted zone is idealized as a cylindrical pipe of uncertain initial diameter at the base of the flaw.** + - *Engineering basis:* The cylindrical-pipe idealization of the Concentrated Leak Erosion Initiation toolbox (see + [Initiation](./06-initiation.mdx#delegated-method-cylindrical-pipe-initiation)); the basis of the default diameter ranges is requested at + marker IN-11 in [Initiation](./06-initiation.mdx#initial-pipe-diameter). + - *Effect on results:* The shear stress rises in proportion to the diameter, so the initial diameter range strongly shapes the initiation curve. + - *When it may not hold:* Where the flaw is better represented as a crack, such as a gap against a structure; the toolbox crack geometries then + apply. +11. **The default critical shear stress is a triangular distribution over a range set by the soil classification.** + - *Engineering basis:* Basis requested — see markers IN-07 and IN-10 in [Initiation](./06-initiation.mdx#critical-shear-stress). + - *Effect on results:* Soils of one classification share one erodibility range regardless of compaction, water content, or dispersivity. + - *When it may not hold:* For dispersive soils, soils whose erodibility differs markedly from their classification's range, or where erosion + testing is available; a site-specific distribution should then be supplied. +12. **The compaction mechanisms are not modeled in clean or nearly clean sands and gravels** — the classes SP-SM, SW, SP, GP-GM, GW, and GP, + with less than 12 percent fines. + - *Engineering basis:* Basis requested — see marker IN-09 in [Initiation](./06-initiation.mdx#critical-shear-stress). + - *Effect on results:* A core of SP-SM, SW, SP, GP-GM, GW, or GP cannot be analyzed by the compaction models. Silty soils (SM, GM, ML) are + analyzed even where they are non-plastic. + - *When it may not hold:* Not applicable within the family; other internal-erosion failure modes address such soils. +13. **The seepage path is a horizontal line at the base of the flaw across the core section, or across the shell when no core is described.** + - *Engineering basis:* The pipe-length measurement of the Concentrated Leak Erosion Initiation toolbox, which measures across the impervious core + of a zoned embankment with very permeable outer zones and across the embankment slopes of a homogeneous one. + - *Effect on results:* The whole net head is lost along the measured length; a shorter length gives a steeper gradient and a higher initiation + probability. + - *When it may not hold:* For sloping cores, flaws not perpendicular to the centerline, or outer zones that are not free draining; a supplied + seepage path length should then be used. +14. **The critical shear stress and the initial pipe diameter are uncertain independently, and each realization holds its values at every pool + level.** + - *Engineering basis:* The probabilistic method of the Concentrated Leak Erosion Initiation toolbox, which samples each from its own + distribution. + - *Effect on results:* The initiation curve never falls where the gradient rises with pool level (it is non-decreasing there). + - *When it may not hold:* Where the two are correlated, for example where a more erodible soil also forms larger initial openings. Under such + positive dependence the ratio of diameter to critical shear stress has a wider spread, so the initiation curve rises over a wider range of + pool levels than the independent treatment shows. + +## Failure Pathway and Progression + +15. **On a crest crack, continuation and roof support share one gate at the top of the first filter the pathway meets.** Above that elevation + continuation is certain and the full-pathway materials govern roof support; at or below it the continuation method and the below-filter + materials apply; a filter reaching the crest with no full-pathway materials given applies at every pool level. + - *Engineering basis:* Basis requested — see marker IN-12 in [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#node-3-continuation). + - *Effect on results:* A crest-crack curve with a filter steps where the pool crosses the filter's top. It steps up when the continuation + probability $P_{ce}$ is smaller than the ratio of full-pathway to below-filter roof support, and down when it is larger (see + [Node 3: Continuation](./07-failure-pathway-and-progression.mdx#node-3-continuation)). + - *When it may not hold:* Where a filter's top varies along the crack, or a second filter lies further downstream on the full pathway; the + first filter's top is used throughout. +16. **On a compaction flaw, the pathway, its filter, its materials, and the walls that reach it are fixed by the flaw elevation.** Continuation, + roof support, and flow limitation are each one value at every pool level. + - *Engineering basis:* Engineering judgment: the flaw is at one elevation, so what it meets does not change with the pool. + - *Effect on results:* Only initiation varies with pool level. + - *When it may not hold:* Where the poorly compacted zone extends over a range of elevations; the flaw elevation should then be chosen as the one + that governs. +17. **Roof support is the minimum roof-support probability across the materials on the pathway.** + - *Engineering basis:* Fell et al. (2024) for the probabilities by soil classification; engineering + judgment for the minimum, since every material along the pathway must hold a roof. + - *Effect on results:* One cohesionless material on the pathway reduces the SRP by up to two orders of magnitude. + - *When it may not hold:* Where the governing material occupies only a short, well-confined length of the pathway. +18. **Crack filling depends only on the embankment type.** + - *Engineering basis:* Fell et al. (2024) ; the meaning of the values is requested at marker IN-13 in + [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#node-5-progression--no-crack-filling). + - *Effect on results:* Only a central-core earth and rockfill embankment receives credit, which halves its SRP. + - *When it may not hold:* Where the gradation of the upstream zone is known to be able, or unable, to fill the pipe. +19. **Flow limitation is the minimum across the applicable upstream elements, and a wall limits flow only below its top.** + - *Engineering basis:* Fell et al. (2024) for the probabilities; the gating conventions are requested at + marker IN-14 in [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#walls-that-stop-below-the-crest). + - *Effect on results:* A modern cutoff wall that reaches the flaw reduces the SRP by three orders of magnitude; above its top, a crest-crack curve + loses that reduction. + - *When it may not hold:* Where a wall is discontinuous, damaged, or does not span the flaw laterally. +20. **The conduit compaction model takes no flow-limitation credit.** + - *Engineering basis:* Basis requested — see marker IN-14 in [Failure Pathway and Progression](./07-failure-pathway-and-progression.mdx#walls-that-stop-below-the-crest). + - *Effect on results:* An upstream cutoff or facing does not reduce the SRP of a conduit flaw (conservative). + - *When it may not hold:* Where an upstream element demonstrably limits flow into the zone around the conduit. + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/09-output-and-interpretation.mdx b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/09-output-and-interpretation.mdx new file mode 100644 index 000000000..7a5f7a679 --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/09-output-and-interpretation.mdx @@ -0,0 +1,130 @@ +--- +title: Output and Interpretation +--- + +import Citation from '@site/src/components/Citation'; +import CitationFootnote from '@site/src/components/CitationFootnote'; +import EquationNoRef from '@site/src/components/EquationNoRef'; +import NavContainer from '@site/src/components/NavContainer'; + + + +# Output and Interpretation + +Every CLE screening model reports the SRP at each evaluated pool level together with the probability of each of the eight nodes, so that the +contribution of every node to the curve can be read directly. The results fall into three groups: values that do not depend on pool level, series +evaluated at each pool level, and the supporting results of the flaw assessment and the initiation analysis. Each model's +own manual describes its flaw-assessment results. + +## Pool-Independent Results + +- **Flaw probability** — Node 1, from the model's flaw assessment (see [Flaw Conventions](./05-flaw-conventions.mdx)); 1.0 when an observation + override applies. +- **Crack-filling probability** — Node 5, the probability of no crack filling for the embankment type. +- **Flow-limitation probability** — Node 6, the minimum across the applicable upstream elements, or 1.0 when none is listed: + - *Crest-crack models* — the value with every listed element applying, which governs at pool levels at or below the top of a wall. + - *Compaction models* — the value after the walls are gated against the flaw elevation, so the walls are excluded when the top of the wall + $z_{tw}$ is below the base of the flaw $z_f$. This is the value used at every pool level. + - *Conduit compaction model* — no flow-limitation result is reported; its node is fixed at 1.0. +- **Roof-support probabilities** — for the crest-crack models, the minimum across the full-pathway materials and the minimum across the + below-filter materials, each reported only where that pathway exists (the full-pathway value is absent when the filter caps the pathway; the + below-filter value is absent when no filter is present). For the compaction models, the single roof-support probability across the embankment + materials. +- **Continuation probability** (compaction models) — Node 3: the continuation method's result when a filter intercepts the pathway, otherwise + 1.0. + +## Pool-Dependent Series + +Every pool-dependent result is a series over the evaluated pool levels — every supplied headwater value and the additional levels distributed +between them (see [Event Tree](./04-event-tree.mdx#headwater-grid)): + +- **Pool levels** — the headwater elevations at which the event tree is evaluated. +- **Tailwater** — the tailwater at each pool level, interpolated between the supplied pairs (see + [Event Tree](./04-event-tree.mdx#headwater-grid)). Where there is no tailwater, the crest-crack models report the entry as blank, while the + compaction models report the pipe-invert elevation $z_f$, the level at which tailwater gives no submergence. +- **Initiation probability** — Node 2 at each pool level. +- **Continuation probability** (crest-crack models) — Node 3 at each pool level: the continuation method's result at or below the filter's top, + and 1.0 above it or where no filter is present. +- **Roof-support probability** (crest-crack models) — Node 4 at each pool level: the below-filter or full-pathway value that governs there. +- **Flow-limitation probability** — Node 6 at each pool level, after any wall has dropped out. For the compaction models it is the same at every + pool level; for the crest-crack models it changes where the pool crosses the top of a wall. The conduit compaction model reports no + flow-limitation series; its Node 6 is fixed at 1.0. +- **Unsuccessful-intervention and breach probabilities** — Nodes 7 and 8, the user-specified values at every pool level. +- **SRP** — the product of the eight nodes at each pool level: the system response curve. + +## Supporting Results + +- **Flaw assessment** — the factor scores, the weighted sum, the raw and final flaw probabilities, whether an override applied, and the flaw + elevation (the bottom of the crack, or the base of the flaw), as documented in each model's manual. +- **Initiation analysis** (compaction models) — at each pool level, the upstream and downstream heads above the pipe invert, the hydraulic + gradient, and the mean and 5th, 50th, and 95th percentiles of the hydraulic shear stress; and the seepage path length. + +## Interpreting the System Response Curve + +The SRP series is the system response curve — the conditional probability of breach by concentrated leak erosion through the modeled flaw as a +function of pool level. Its shape differs between the two groups. + +**Crest-crack models.** The curve is negligible at and below the bottom of the crack. It steps up as soon as the pool rises above it, provided +every other node is non-zero there: a negligible flaw probability (an inapplicable mechanism, or a factor or threshold forcing it to zero), or a +user-supplied unsuccessful-intervention or breach probability of 0, keeps the curve negligible at every pool level. An observation override takes +precedence over the threshold rule, so a reported observation restores a flaw probability of 1.0 (see +[Flaw Conventions](./05-flaw-conventions.mdx#flaw-probability-from-the-weighted-sum)). Above that level the curve is piecewise +constant: with initiation equal to 1.0, the SRP at each pool level is the product of the remaining nodes, + + + +with $P_{cont}$, $P_{roof}$, and $P_{nl}$ taking the values that govern at that pool level. Further steps occur where the pool crosses the top of +a filter (continuation becomes certain and the full-pathway roof materials take over, so the curve can step up or down; see +[Node 3: Continuation](./07-failure-pathway-and-progression.mdx#node-3-continuation)) and where it crosses the top of a wall (the wall's +flow-limitation credit is lost). Each step location can be read from the node series. + +**Compaction models.** The curve is negligible at and below the base of the flaw and rises with pool level as the hydraulic gradient through the +pipe grows, following the initiation probability. The other seven nodes are constant, so the curve approaches the product of those nodes as the +initiation probability approaches 1.0. A rising tailwater that reduces the net head across the pipe can flatten or lower the curve over that range. + +## Reading the Node Breakdown + +Because the SRP is a product, the node with the lowest probability at a pool level limits the curve there, and the breakdown shows which it is. +Typical readings: + +- **A low flaw probability** places the whole curve low; it reflects the factor scores, and the flaw assessment results show which factors drive + it. +- **A filter on the pathway** lowers the curve by the continuation probability — on a crest crack only up to the filter's top. +- **A cohesionless material on the pathway** lowers the curve by its roof-support probability, by up to two orders of magnitude. +- **An effective upstream element** lowers the curve by its flow-limitation probability, but only where it applies. +- **For the compaction models**, the pool level at which the curve begins to rise, and how fast it rises, are set by the critical shear stress and + initial pipe diameter. A curve that levels off below the product of the other nodes reflects realizations that do not initiate within the + evaluated pool range. Such realizations may never initiate at any physically possible pool, because the crest limits the largest gradient the + pipe can carry, so the plateau can be real. Extending the headwater series shows whether the curve rises further, but only up to the crest or + the maximum credible pool. + +## Checking the Result + +In the crest-crack models, the curve can be reproduced exactly from the reported node values, since initiation, roof support, crack filling, and +flow limitation are deterministic. The continuation probability $P_{ce}$ itself may be a Monte Carlo estimate, when the continuation method +evaluates the filter by its sampling route through the USCS classes of the base and filter soils. In the compaction models, the +initiation probability at each pool level is a Monte Carlo point estimate, the fraction of realizations that initiate, and no confidence bounds +are reported with it; small irregularities between neighboring pool levels can reflect sampling rather than the physics. The following checks +apply to every result: + +- **Upper bound** — in the compaction models, the SRP at every pool level is no larger than the product of the seven nodes that do not vary with + pool level, and approaches it only as initiation approaches 1.0. In the crest-crack models, the SRP above the bottom of the crack equals the + product of the nodes that govern at that pool level. +- **Step locations** — in the crest-crack models, steps of the curve can occur only at the bottom of the crack $z_{bc}$, the top of the filter + $z_{tf}$, and the top of a wall $z_{tw}$ (to within one interval of the headwater grid), and at a filter or wall top only where it lies above + $z_{bc}$. A filter or wall top at or below the bottom of the crack produces no visible step. +- **Lower end** — the curve is negligible at and below the bottom of the crack $z_{bc}$ (crest-crack models) or the base of the flaw $z_f$ + (compaction models). + +A curve that fails one of these checks can point to an input that does not describe the intended embankment, such as a flaw elevation, filter top, +or wall top on a different datum from the headwater series. + +## Screening Results Relative to Detailed Analysis + +The CLE screening models are screening tools. Their results indicate whether a concentrated leak erosion mechanism warrants detailed analysis; +they are not substitutes for a detailed assessment. The screening defaults — the critical shear stress and initial pipe diameter ranges, the +flaw-elevation heuristics, the certain initiation of an open crest crack, and the tabulated progression probabilities — are chosen to avoid false +negatives. Where a screening result is significant to a decision, the governing nodes should be re-examined with site-specific information, and the +detailed toolbox methods, the methods of Fell et al. (2024) , and the judgment of the risk team should replace the screening values. + + diff --git a/docs/web-applications/system-response/cle-common-methods-screening/v1.0/_input-needed.md b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/_input-needed.md new file mode 100644 index 000000000..e45d379d3 --- /dev/null +++ b/docs/web-applications/system-response/cle-common-methods-screening/v1.0/_input-needed.md @@ -0,0 +1,249 @@ +# Input Needed — Concentrated Leak Erosion Common Methods Screening Technical Reference Manual (v1.0) + +Answer inline and return this file (or paste its contents into the conversation). Leave an +Answer blank to keep that marker in place. Do not renumber — the IDs tie each entry to its +admonition in the document. + +--- + +## IN-01 — Applicability — `03-input-parameters.mdx` + +**Regarding:** The length unit the CLE family is intended to be used in. Several +mechanism-specific relationships carry dimensional values with no unit conversion (the +desiccation and frost-cracking crack depths are tabulated in feet; the compaction models band +the flaw base width at fixed numerical thresholds), only the cross-valley settlement model +converts its crack-depth relationship between feet and meters, and the shear stress +relationship uses the unit weight of water in pounds per cubic foot. + +**Needed:** The intended unit system for the family, and whether the statement belongs in this +manual or in each model's manual. + +**Proposed (unconfirmed):** The family is intended for US customary units (elevations and +lengths in feet); a user working in SI units converts before running any model other than the +cross-valley settlement model. + +**Answer:** + +--- + +## IN-02 — Definition — `03-input-parameters.mdx` + +**Regarding:** The upstream-element categories "modern cutoff wall" and "non-cementitious +wall", which are listed without a definition alongside the soil-cement-bentonite and column-wall +categories. + +**Needed:** What each category covers (construction type, era, or material), or the table in +Fell et al. (2024) that defines them. + +**Answer:** + +--- + +## IN-03 — Engineering Basis — `05-flaw-conventions.mdx` + +**Regarding:** The conversion of the risk-factor/likelihood-factor weighted sum to a flaw +probability by interpolating the logarithm of probability linearly between tabulated points, +holding the end value beyond the tabulated range, and (in several models) anchoring the lowest +tabulated point at a negligibly small probability. + +**Needed:** Whether Fell et al. (2024) presents these relationships as curves to be +interpolated this way (with the figure or table), or the engineering-judgment attribution. + +**Proposed (unconfirmed):** The tabulated probabilities span several orders of magnitude, so +interpolation in the logarithm of probability is the natural reading of a relationship plotted +on a logarithmic probability axis; the negligible anchor reproduces a relationship that falls +toward zero at the lowest attainable score. + +**Answer:** + +--- + +## IN-04 — Engineering Basis — `05-flaw-conventions.mdx` + +**Regarding:** The flaw-continuity stage applied by the core, wall or cliff, and foundation or +abutment compaction models, and its shared curve (likelihood factor 1, 2, 3, 4 → negligible, +1 × 10⁻⁴, 1 × 10⁻³, 1 × 10⁻¹), keyed to a different compaction factor in each model. + +**Needed:** The physical meaning of the flaw-continuity probability, its source (or the +engineering-judgment attribution), and confirmation that one curve is intended for all three +models. + +**Answer:** + +--- + +## IN-05 — Engineering Basis — `05-flaw-conventions.mdx` + +**Regarding:** The normal-pool heuristic placing the base of a core or abutment compaction flaw +one third of the way from the normal pool elevation to the crest. + +**Needed:** Literature source or engineering-judgment attribution, including why the normal pool +and not the base of the embankment is the lower reference. + +**Proposed (unconfirmed):** A screening convention that places the flaw within the band of pools +the embankment experiences, so that initiation is evaluated above normal pool where loading is +frequent. + +**Answer:** + +--- + +## IN-06 — Engineering Basis — `06-initiation.mdx` + +**Regarding:** The initiation probability of 1.0 applied by the crest-crack models once the pool +rises above the bottom of an open crest crack (0 at or below it). The model records two +different attributions: a conservative screening convention that is not an explicit finding of +Pierson (2023), and an attribution of the certain initiation to Pierson (2023). + +**Needed:** Which attribution is intended. If Pierson (2023) is to be cited, its full +bibliographic details (authors, title, publisher or journal, report number) for the +bibliography. + +**Proposed (unconfirmed):** An open transverse crack at the crest is wide enough that the shear +stress of any flow through it exceeds the critical shear stress of the crack walls, so +initiation is taken as certain once water can enter the crack — a conservative screening +posture under which a site needing a lower value adjusts the SRP. + +**Answer:** + +--- + +## IN-07 — Engineering Basis — `06-initiation.mdx` + +**Regarding:** The default critical shear stress ranges by soil classification shared by the +four compaction models (lb/ft²: SM, SC, ML 0.02089 / 0.04177 / 0.1044; GM 0.021 / 0.08354 / +0.2089; CL-ML 0.04177 / 0.08354 / 0.2089; GC 0.04177 / 0.08354 / 0.418; CL, MH 0.04177 / +0.1044 / 0.4177; CH 0.2089 / 0.5221 / 1.044). + +**Needed:** Literature source with full bibliographic details for the bibliography, or the +engineering-judgment attribution. + +**Proposed (unconfirmed):** The values are conversions of round values in pascals (1, 2, 4, 5, +10, 20, 25, 50 Pa), suggesting a source table published in SI units. + +**Answer:** + +--- + +## IN-08 — Decision — `06-initiation.mdx` + +**Regarding:** Two entries of the critical shear stress table that do not follow the pascal +conversion of the rest: the GM low value 0.021 lb/ft² (1 Pa = 0.02089 lb/ft²) and the GC high +value 0.418 lb/ft² (20 Pa = 0.4177 lb/ft²). + +**Needed:** Whether these values are intended as printed or should match the converted values. + +**Answer:** + +--- + +## IN-09 — Engineering Basis — `06-initiation.mdx` + +**Regarding:** The exclusion of the six clean and nearly clean sand and gravel classes (SP-SM, +SW, SP, GP-GM, GW, GP; less than 12 percent fines) as the +core soil classification of all four compaction models. + +**Needed:** Engineering basis or engineering-judgment attribution. + +**Proposed (unconfirmed):** A cohesionless soil cannot sustain an open pipe or crack, so a +compaction flaw in it would collapse rather than carry a concentrated leak; such soils are +addressed by other internal-erosion failure modes. + +**Answer:** + +--- + +## IN-10 — Engineering Basis — `06-initiation.mdx` + +**Regarding:** The triangular form of the default critical shear stress distribution over the +tabulated low, most likely, and high values. + +**Needed:** Engineering basis or engineering-judgment attribution. + +**Proposed (unconfirmed):** The triangular form follows the probabilistic input of the +Concentrated Leak Erosion Initiation toolbox, which represents the critical shear stress and the +initial pipe diameter by triangular distributions over their minimum, most likely, and maximum +values. + +**Answer:** + +--- + +## IN-11 — Engineering Basis — `06-initiation.mdx` + +**Regarding:** The default initial pipe diameter ranges (triangular, mm): core compaction 2 / 3 +/ 5; conduit compaction 5 / 10 / 12.5; wall or cliff compaction 5 / 5 / 10; foundation or +abutment compaction 2 / 3 / 5. + +**Needed:** Literature source or engineering-judgment attribution, including why the conduit +range is larger than the core range and why the wall or cliff range has its most likely value at +its low value. + +**Answer:** + +--- + +## IN-12 — Engineering Basis — `07-failure-pathway-and-progression.mdx` + +**Regarding:** The crest-crack pathway convention: (1) continuation and roof support share one +gate at the top of the first filter the pathway meets, with certain continuation above it; (2) +the below-filter roof materials are taken at the bottom-of-crack elevation; (3) a filter whose +top reaches the crest, with no full-pathway materials given, caps the pathway at every pool +level. + +**Needed:** Literature source or engineering-judgment attribution. + +**Proposed (unconfirmed):** A crest crack is open over its full depth, so a pool above the +filter's top drives flow through the upper part of the crack over the filter, where nothing +filters the exit; below the filter's top, all flow through the crack reaches the filter. + +**Answer:** + +--- + +## IN-13 — Definition — `07-failure-pathway-and-progression.mdx` + +**Regarding:** The meaning of the crack-filling values by embankment type (homogeneous earthfill +1.0, zoned earthfill 1.0, central-core earth and rockfill 0.5). The source table is described +as giving the probability of crack filling, while the event tree multiplies the value directly +as the probability of no crack filling. + +**Needed:** Which reading the source table supports, and its table number in Fell et al. (2024). + +**Proposed (unconfirmed):** The value is the probability that crack filling does not stop the +erosion — an earthfill embankment has no upstream zone of material able to fill the pipe, so no +credit is taken, while the upstream zone of a central-core earth and rockfill embankment is +credited with a one-in-two chance of filling it. + +**Answer:** + +--- + +## IN-14 — Engineering Basis — `07-failure-pathway-and-progression.mdx` + +**Regarding:** The flow-limitation conventions: (1) the minimum across several upstream elements +governs; (2) a wall stops limiting flow above its top — per pool level for a crest crack, and +against the base of the flaw for a compaction flaw — while an upstream facing element applies at +every pool level; (3) the conduit compaction model takes no flow-limitation credit. + +**Needed:** Literature source (including the table in Fell et al. (2024) that gives the +flow-limitation probabilities) or the engineering-judgment attribution. + +**Answer:** + +--- + +## IN-15 — Definition — `03-input-parameters.mdx` + +**Regarding:** The elevation, or range of elevations, at which the full-pathway roof materials +of the crest-crack models are taken. The below-filter materials are taken at the +bottom-of-crack elevation; the full-pathway materials are described only as every non-filter +material from the upstream face to the downstream face. + +**Needed:** The elevation or elevation range at which the user should cut the section when +listing the full-pathway materials (for example, the bottom of the crack, the top of the +filter, or the whole depth of the crack above the filter's top). + +**Answer:** + +--- diff --git a/src/docConfig.js b/src/docConfig.js index 94aa4cf81..7593fb7df 100644 --- a/src/docConfig.js +++ b/src/docConfig.js @@ -345,7 +345,7 @@ const docs = [ software: 'system-response', doc_location: 'web-applications/system-response/cle-common-methods-screening', doc_name: 'Concentrated Leak Erosion Common Methods Screening Technical Reference Manual', - active: false, + active: true, draft: true, }, { diff --git a/static/bibliographies/web-applications/system-response/cle-common-methods-screening/v1.0/bib.json b/static/bibliographies/web-applications/system-response/cle-common-methods-screening/v1.0/bib.json new file mode 100644 index 000000000..e5c80cf3b --- /dev/null +++ b/static/bibliographies/web-applications/system-response/cle-common-methods-screening/v1.0/bib.json @@ -0,0 +1,17 @@ +[ + { + "citationKey": "Fell2024", + "entryType": "report", + "author": [ + "R. Fell", + "M.A. Foster", + "J. Cyganiewicz", + "R.R. Davidson", + "G.L. Sills" + ], + "title": "Methods for Estimating Probability of Failure of Embankment Dams by Internal Erosion and Piping (The Piping Toolbox 2024)", + "year": 2024, + "institution": "University of New South Wales, School of Civil and Environmental Engineering", + "report": "Report No. 470" + } +] From c85cb81a3bdb9dcefa5026a578ffebb7622262c8 Mon Sep 17 00:00:00 2001 From: Adam Gohs Date: Fri, 16 Oct 2026 09:50:49 -0400 Subject: [PATCH 2/2] Set the manual's sidebar and breadcrumb name --- src/docConfig.js | 1 + 1 file changed, 1 insertion(+) diff --git a/src/docConfig.js b/src/docConfig.js index 7593fb7df..7af9ab6ce 100644 --- a/src/docConfig.js +++ b/src/docConfig.js @@ -345,6 +345,7 @@ const docs = [ software: 'system-response', doc_location: 'web-applications/system-response/cle-common-methods-screening', doc_name: 'Concentrated Leak Erosion Common Methods Screening Technical Reference Manual', + sidebar_label: 'Concentrated Leak Erosion Common Methods Screening', active: true, draft: true, },