Dependency Cut-Set Review¶
Analytical review — instantiates Cascade Pathway Management
Analyzes the dependency structure to find the minimal set of links whose removal isolates harm, telling you exactly where a breakpoint or safety gate should sit.
A Dependency Cut-Set Review is the analysis that answers "if we could cut a few links, which ones?" Given a map of the system's dependencies, it searches for a cut set — a set of links whose removal disconnects the source of harm from what must be protected — and looks specifically for the minimal such set, the smallest number of cuts that still isolates the danger. Its defining property is that it is prescriptive and selective: it does not describe the whole web and it does not install anything at runtime; it reasons over the structure to nominate the few high-value places where a breakpoint or safety gate should go, and to prove that cutting there actually severs the harmful path. It produces a recommendation — a target list — not a diagram and not a device.
Example¶
A regional power utility must ensure that a fault on one transmission line cannot black out the whole grid. Its dependency map is huge — hundreds of lines, buses, and substations — and it cannot install a protective breaker everywhere. So the reliability team runs a cut-set review. They take the harmful outcome to prevent ("the western load center loses all supply") and search the network for the minimal sets of lines that, if opened, would isolate a fault before it can propagate across the interconnection.
The analysis surfaces that a particular pair of tie-lines forms a critical cut: sever those two and a cascading overload originating in the north is walled off from the western center, while every other path stays energized. It also finds a near-miss — a three-line cut that would work but strands a hospital feeder, so it is rejected in favor of the two-line option. The team runs the classic N-1 contingency test against their proposed cuts to confirm that losing any single element beyond the cut still leaves the system stable.[1] The deliverable is a short, ranked list: place a protective device on these two lines, and here is the proof they form a valid isolating cut. The review installs nothing — it tells the breaker program exactly where to aim.
How it works¶
- Fix the source and the protected set. The review names the harm to contain and the assets to keep connected; a cut set only makes sense relative to "keep this safe from that."
- Search for isolating cuts, ranked by minimality. It enumerates sets of links whose removal disconnects source from protected asset, preferring the smallest and least costly — fewer cuts mean less lost connectivity.
- Score each candidate cut's collateral. Every cut severs useful flow too; the review weighs each candidate's containment value against what legitimately depended on the severed links, discarding cuts that strand something critical.
- Recommend placement, prove sufficiency. It hands off a ranked list of where to install breakpoints or safety gates, with the demonstration that cutting there genuinely isolates the harm.
Tuning parameters¶
- Cut-size ceiling — the maximum number of links a proposed cut may contain. A tight ceiling forces elegant single-choke solutions but may find none; a loose ceiling always finds a cut but may recommend carving the system into pieces.
- Collateral weighting — how heavily to penalize a cut for the useful flow it destroys. Heavy weighting protects functional connectivity but may leave no acceptable cut; light weighting isolates aggressively at the cost of amputation.
- Contingency depth — whether cuts must survive one failure beyond themselves (N-1), two (N-2), or none. Deeper contingency yields more robust placements but sharply narrows viable options.
- Recompute trigger — whether the review reruns on every structural change or only periodically. Frequent reruns keep recommendations valid as the map evolves; infrequent ones risk placing devices on a cut that has since stopped being a cut.
When it helps, and when it misleads¶
Its strength is turning an intractable "where do we put our limited protective devices?" into a proven, ranked shortlist — it finds the few chokepoints that isolate the most harm for the least lost connectivity, and it shows its work. It is the analysis that makes a circuit-breaker or gate program targeted rather than scattershot.
Its failure mode is that the cut is only valid against the map it reasoned over. If the dependency map omits an informal bypass, the recommended cut isolates nothing — harm simply routes around it through a link the review never saw, the archetype's hidden bypass failure wearing an analytical mask. And a cut set validated last year can be silently invalidated by rewiring, so a device placed on a former chokepoint guards a path that no longer exists. The classic misuse is trusting a cut-set recommendation as permanent truth and never revalidating it. The guarding discipline is to rerun the review against a freshly verified map after any structural change, and to stress the proposed cuts against the possibility of an unmapped parallel path.
How it implements the components¶
cascade_breakpoint_set— its output is the breakpoint set: the specific, minimal collection of links where cutting isolates the harm, proven sufficient.counter_cascade_safety_gate— where a full cut costs too much, it recommends a safety gate on the same chokepoint instead, nominating where a conditional isolating control belongs.
This is analysis, not machinery and not description. It does not enumerate every link or draw the web — the exhaustive transfer_link_inventory and propagation_pathway_map belong to its nearest twin, Cascade Dependency Graph, which it consumes as input. And it does not execute the cut at runtime or hold a response_authority_map; firing the breakpoint it nominates is Cascade Circuit Breaker. The graph maps the wiring, this review picks where to cut it, and the breaker does the cutting.
Related¶
- Instantiates: Cascade Pathway Management — the review is the archetype's targeting step for where to place breakpoints and gates.
- Consumes: Cascade Dependency Graph — it reasons over the graph's link inventory to find cuts.
- Sibling mechanisms: Cascade Dependency Graph · Cascade Circuit Breaker · Rate-Limit or Quarantine Gate · Consequence Cascade Workshop · Domino Tabletop Exercise · Leading-Link Indicator Dashboard · Propagation Simulation or Fault Injection · Beneficial Cascade Seeding Plan
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Dependency Cut-Set Review operates as a computation, comparison, model, or analytic representation used to infer, estimate, or choose because it analyzes the dependency structure to find the minimal set of links whose removal isolates harm, telling you exactly where a breakpoint or safety gate should sit.
Independent corroboration: The frozen evidence defines Dependency Cut-Set Review as 'Analyzes the dependency structure to find the minimal set of links whose removal isolates harm, telling you exactly where a breakpoint or safety gate should sit', so its operative form is Analysis, Modeling & Optimization.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Engineering & Design
Origin pattern: Single lineage
Present-day reach: Multi-domain
Rationale: Reliability and safety engineering cohered minimal cut-set analysis for identifying the smallest link set whose removal isolates a hazard.
Related originating lineages:
- Mathematics — Graph theory supplied edge cuts and minimal separators.
- Operations Research — Network interdiction and reliability optimization supplied selective cut placement.
Review resolution: Reliability and safety engineering cohered minimal cut-set analysis for identifying the smallest link set whose removal isolates a hazard. Graph cuts and operations-research network analysis are genuine methodological antecedents to engineering reliability cut-set review.
Review outcome: Reconciled after independent review; high confidence.
References¶
[1] Wood, A. J., Wollenberg, B. F., & Sheblé, G. B. Power Generation, Operation, and Control (3rd ed.). John Wiley & Sons (2013). Defines the N−1 security criterion as requiring the power system to remain secure after the loss of any single relevant element. registry ↩