A13. Settlement Layout Design Elements

This appendix expands the eight CLS elements. For each sub-element you’ll find: purpose (what it promotes or prevents), a how to estimate outline (light math/procedure), and what good / bad typically looks like. All sub-elements are normalized to 0–1 (higher is better) and averaged within their element; elements are then combined with the headline weights.

1) Safety & Resilience (weight 0.33)

Code	Sub-element	Purpose (promote / prevent)	How to estimate (outline)	“Good” looks like	“Bad” looks like
S1	Egress coverage	Promote fast access to a refuge; prevent entrapment.	Shortest-time from each occupied node to any haven via pressurized links; compare to target (e.g., 60 s). Population-weighted share ≤ target.	≥ 90% of people ≤ 60 s to haven.	Large pockets > 60 s; dead-ends near far corners.
S2	Route redundancy	Promote 2 independent ways to safety; prevent single-route dependency.	Count edge-disjoint paths from each node to a haven; cap at 2; average across nodes.	Almost all nodes have 2 disjoint routes.	Many nodes rely on a single corridor.
S3a	Max sealed compartment size	Limit consequences of breach/fire.	“Close” bulkheads; find largest isolated compartment area; compare to cap (e.g., ≤ 1,500 m²).	Largest ≤ cap with margin.	One mega-compartment spans a whole wing.
S3b	Compartment density	Balance “too few” vs “too many” pressure zones.	Number of sealed compartments per 10,000 m² vs a target derived from preferred compartment size (e.g., 1,200–1,500 m²).	Near the middle band; neither sparse nor fussy.	Either one huge zone or dozens of tiny cells.
S4	Dirty↔Clean segregation	Maintain hygiene & pressure discipline; prevent contaminant shortcuts.	For every dirty→clean node pair, check that all shortest paths include ≥ 1 bulkhead/interlock. Score = compliant pairs share.	All shortest paths interlocked.	Direct dirty→clean shortcuts abound.
S5	Shielding minima	Protect sleep/med spaces; prevent under-shielding.	HAB/MED nodes with areal density ≥ target (e.g., 10–20 g/cm² WE), area-weighted share.	~100% HAB/MED compliant.	Significant HAB/MED below target.
S6	Network robustness	Promote graceful degradation; prevent cascade failure.	Proxy by algebraic connectivity of the circulation graph or pass/fail on single-edge-removal egress drill.	Egress still passes when any one link fails.	One cut isolates large populations.

Design moves that lift Safety: close loops (people & utilities), regular sector bulkheads, distributed havens, enforce dirty/clean interlocks on shortest paths.

2) Flow & Operations (weight 0.18)

Code	Sub-element	Purpose	How to estimate	Good	Bad
F1	Travel efficiency	Shorten everyday trips; reduce fatigue/time loss.	OD matrix for key pairs (HAB↔MESS, HAB↔MED, WRK↔STOR, etc.); compute weighted mean travel time; compare to target (e.g., 120 s).	Most key trips feel “nearby”.	Long criss-crossing walks.
F2	Chokepoints	Avoid overloaded corridors/doors.	Assign OD flows along shortest paths; compute utilization vs capacity per link/door; take the max.	No element regularly > 80% at peak.	One spine repeatedly saturates.
F3	Operational path diversity	Keep ops running during blockages.	Share of important OD pairs with ≥ 2 disjoint paths whose length ≤ α×shortest (e.g., α = 1.4).	Two workable routes almost everywhere.	Alternatives are long detours or nonexistent.
F4	Bulkhead-crossing friction	Balance safety gates vs day-to-day flow.	OD-weighted mean pressure-door crossings per shortest path; target a moderate band.	~0.3 crossings/trip typical.	Every short trip hits multiple bulkheads.

Design moves: add cross-ties to shorten detours; don’t place bulkheads on every doorway—use sector lines.

3) Plan & Envelope (weight 0.08)

Code	Sub-element	Purpose	How to estimate	Good	Bad
P1	Link-length density	Penalize gratuitous corridor sprawl; reward clean plans.	Total link length per 1,000 m²; score peaks in a sweet-spot band (not too little, not too much).	Near the center of the band.	Over-meshed spaghetti or starved single spine.
P2	Envelope compactness	Reduce surface/volume (thermal, mass).	External surface ÷ pressurized volume vs a reference target.	Compact perimeter & roof.	Long skinny envelope with high S/V.
P3	Site coverage balance	Leave room to breathe inside site hull.	Pressurized footprint ÷ convex-hull area; mid-band rewarded.	~70% coverage.	> 90% coverage (wall-to-wall).
P4	Internal reserve ratio	Reserve real lots for future domes/industry.	Share of hull set aside as platted reserves (each ≥ area & width minima).	~15% reserves platted.	No reserves (or tiny slivers).
P5	Reserve lot quality	Ensure reserves are usable options.	Lot passes 4 tests: min width, ROW access, utility stubs, no critical trunk bisecting. Score = compliant lots share.	Most reserves “plug-and-play”.	Reserves stranded or bisected by utilities.

Design moves: target one or two future “park-dome” lots; keep internal boulevards/alleys that touch reserves.

4) Mass Efficiency (proxy) (weight 0.08)

Code	Sub-element	Purpose	How to estimate	Good	Bad
M1	Joints per node (typed)	Limit interfaces (heavy, leaky, failure-prone).	Simple joints/node in a sweet-spot band; pressure joints weighted more heavily.	~2–2.5 joints/node; few pressure doors.	High joint counts; many pressure doors.
M2	Geometry simplicity /100 m	Fewer bends & complex junctions = lighter, easier.	Counts per 100 m: 90° bends, other kinks, T- and 4-way junctions vs caps.	Long straights; T’s over 4-ways.	4-ways proliferate; snaky corridors.
M3	Curved fraction	Penalize bespoke curved link pieces.	Curved length ÷ total link length; cap low.	≤ 5–10%.	Many custom curves.
M4	Civil works penalty	Keep bridges/underpasses modest.	Bridge/underpass length per 1,000 m²; normalize vs lenient caps.	Minimal underpasses; few bridges.	Frequent underpasses; long engineered works.

Design moves: standardize straight pieces; favor T-junction kits; avoid four-way intersections unless absolutely needed.

5) Construction & Maintenance (weight 0.08)

Code	Sub-element	Purpose	How to estimate	Good	Bad
C1	Service access	Ensure crews can reach utilities.	Fraction of link length with ≥ 1 m clear both sides of trunks.	High share with clearances.	Utilities buried behind tight walls.
C2	Isolation spacing	Make faults small & fixable.	95th-percentile distance to nearest valve/bulkhead; compare to target (e.g., ≤ 30 m).	Short reach to isolation points.	Long runs without isolation.
C3	Junction serviceability	Safe lay-down & turn-outs.	Share of junctions with alcove/maintenance bay.	Most junctions serviceable.	Service carts block corridors.
C4	Link SKU standardization	Fewer part types to stock & fabricate.	Distinct link piece types vs target set (~6: straight, 90°, T, cross, door, taper).	≤ target types.	Many bespoke parts.
C5	Node SKU standardization	Repeatable bays; easier re-use.	Distinct node bay types vs target (e.g., ≤ 6 of m×n multiples).	Small family of bay types.	Every room a one-off size.
C6	Modular link fit	Prefer base length + small shims.	For each straight link, mismatch to base length L₀ within tolerance gets full credit; large remainders are penalized.	Most links “snap” to L₀±10%.	Many odd leftover lengths.
C7	Heavy-access coverage	Bring cranes/rovers to the work.	Share of nodes reachable via surface access graph meeting width/clearance/radius/grade specs.	≥ 85–95% nodes reachable.	Interior unreachable without cuts.
C8	Heavy-access distance	Keep service drives short.	Median perimeter-to-node drive distance along feasible routes.	≤ 150–180 m.	Long detours through the site.
C9	Barrier-crossing burden	Avoid access that relies on many bridges/underpasses.	Mean # crossings on least-cost access route; score mid-band.	≈ 0.3 per trip.	Multiple engineered crossings per trip.

Design moves: plat a service-alley lattice; keep underpasses rare; put isolation valves where people can actually reach them.

6) Expandability (weight 0.08)

Code	Sub-element	Purpose	How to estimate	Good	Bad
X1	Attachment edge density	Many places to grow.	Candidate connection faces per 1,000 m² vs target.	High, evenly distributed.	Few attach points or all on one side.
X2	Live expansion	Build without shutting down.	Share of faces that can be built out while keeping services/live traffic.	≥ 50–80% live.	Most growth requires outages.
X3	Graph transformability (next IL)	Be “close” to the next topology.	Cycle density vs target (spine→ring; ring→mesh) plus pre-stubbed cross-links credit.	Reserved cross-ties; cycles near target.	Beautiful now, but no path to ring/mesh.
X4	Upgrade disruption	Keep offline area small.	Minimal concurrent depressurized area required to reach next IL; compare to target (e.g., ≤ 10–15%).	Low outage footprint.	A third of the complex offline to upgrade.
X5	Reserved ROW coverage	Guarantee space for future ties/alleys.	Interior blocks touched by platted ROW / total interior blocks.	ROW grid touches most blocks.	ROWs missing; future ties blocked.
X6	Utility readiness	Pre-stub for the next loop(s).	Share of primary runs with spare capacity and sectionalizer/valve stubs at spacing.	Most primaries ready to loop.	Rework required to add loops.
X7	Mergeability index	Combine bays later into big lots.	Average # demountable bays to remove to yield ≥ lot-min area with ROW & width; invert to a 0–1 score.	Few bays to clear a park/industry lot.	Walls & loads make recombination painful.

Design moves: draw the future streets/utilities now; use demountable bay walls; size door frames to upgrade to bulkheads.

7) Habitability (weight 0.09)

Code	Sub-element	Purpose	How to estimate	Good	Bad
H1	Commons accessibility	Keep life near food/social spaces.	HAB population within target time (e.g., 90 s) to MESS/REC.	Most sleepers near commons.	Long walks to eat or meet.
H2	Adjacency buffering	Protect quiet rooms from noise/vibration.	Penalize direct doors between HAB/MED and WRK/GAR; count buffered cases positively.	Few/no bad adjacencies.	HAB backing onto workshops.
H3	Daylight/biophilia proxy	Support circadian health & morale.	Share of HAB/COMMONS area with daylight/lightwells or certified circadian program.	High share with credible lighting spec.	Windowless warren; poor light program.

Design moves: distribute commons; put buffers (storage, corridors) between

noisy and quiet; plan lightwells or circadian lighting.

8) Environmental & Energy (weight 0.08)

Code	Sub-element	Purpose	How to estimate	Good	Bad
N1	Thermal zoning	Keep heat where it belongs.	High-heat nodes (WRK, GH) bulkhead-separated from HAB and on independent HVAC loops; share compliant.	Most hot uses isolated & looped.	Workshops dumping heat into living wings.
N2	Utility loop redundancy	Survive faults without blackout.	For power/air/water, share of nodes with ≥ 2 independent feeds (≥ 3 for criticals at higher ILs); average over utilities.	Dual-fed almost everywhere; tri-fed criticals.	Many single-fed nodes; co-routed “independent” loops.
N3	External interface discipline	Sensible number of suit/rover airlocks.	Airlocks per 10k m² vs target band (too many = leakage & complexity; too few = ops pain).	Near target with good siting.	One per room (leaky) or just one for all (queue hell).

Design moves: independent utility galleries by district; ring mains with sectionalizers; rational airlock siting strategy.

Global Adjustments (applied after combining elements)

Path-Dependence Risk (PDR): small deductions (−0…−5) if design hardens itself into a local maximum: long cul-de-sacs, co-routed “independent” utilities, blocked ROWs, non-upgradeable doors, off-grid bays.

Over-pack penalty: −3 if site coverage κ ≥ 0.90; −6 if κ ≥ 0.95 (no room left inside the hull).

Option-value credit: roughed-in features (capped cross-links, empty conduits, platted reserves) earn half-credit in their respective subs so foresight is rewarded without full immediate complexity.

Using the Deep-Dive in Practice

As a checklist: Each sub-element is a yes/no-leaning assessment plus a short count or distance. You can score most of CLS from a plan set and a one-page data sheet.

As a design guide: If an element scores low, the table tells you what to build next: close a loop (S1/S2/F3), platt a reserve (P4/P5), add stubs & sectionalizers (X5/X6), insert a service alley (C7–C9), etc.

Across eras: Early (IL-0..2) spines should bias C-pillar (access, isolation) and X-pillar (future loops). Mid (IL-3..4) rings focus on S, F, and P-reserves. Late (IL-5..7) meshes push S/F high while keeping Mass and C under control with disciplined geometry and service-alley lattices.