agentleFS
equinorequinor/neqsim/.github/skills/neqsim-firewater-deluge-design/SKILL.md

neqsim-firewater-deluge-design

Skill167 starsChanged 7 months agoApache-2.0View on GitHub

Fire-water and deluge coverage design per NORSOK S-001 / ISO 13702 / NFPA 15: area versus dedicated object demand, nozzle-net sizing from flow and spacing, fire monitors with wind drift, hydraulic feasibility, active-versus-passive substitution. USE WHEN: a task asks if an area has adequate deluge coverage, how many nozzles are needed, or whether monitors or PFP can replace fire water.

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill for public, educational fire-water coverage screening.

    line 16

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Fire-water and Deluge Coverage Design

Use this skill for public, educational fire-water coverage screening. It turns a protected area and an application rate into a water demand, a nozzle count and a grid pitch, and it tests the two alternatives that get proposed whenever a fixed deluge net is inconvenient: protecting the individual items instead of the area, and using fire monitors instead of a fixed system.

When to Use

  • A verification finding says an area lacks active fire protection and the demand must be scoped.
  • A modification adds hydrocarbon-bearing equipment to an area and the fire-water demand changes.
  • Somebody proposes fire monitors, or passive fire protection, in place of deluge.
  • You need a first nozzle count and grid pitch before a hydraulic network calculation.

Inputs

  • protected_area_m2: plan area for general area coverage, m².
  • area_rate_lpm_per_m2: application rate, (l/min)/m². Public minima are 10 for process areas and equipment surfaces and 20 for wellhead areas and riser balconies.
  • objects: tuple of (tag, surface_area_m2, rate_lpm_per_m2) for dedicated protection.
  • duration_min, foam_concentrate_percent, simultaneous_area_factor.
  • Nozzle: nozzle_k_lpm_per_sqrt_bar, nozzle_min_pressure_barg, max_spacing_m, operating_pressure_barg, coverage_efficiency.
  • Monitor: monitor_count, monitor_flow_lpm, wind_speed_m_s, fall_height_m, characteristic_dimension_m, line_of_sight_obstructed.

Outputs

  • area_demand_lpm, object_demand_lpm, total_demand_lpm, total_demand_m3_per_h, water_volume_m3, foam_concentrate_m3, object_to_area_ratio.
  • nozzle_count_from_flow, nozzle_count_from_coverage, nozzle_count, governing_criterion, grid_spacing_m, delivered_density_lpm_per_m2, pressure_adequate, density_met.
  • nominal_density_lpm_per_m2, drift_displacement_m, wind_coverage_fraction, effective_density_lpm_per_m2, verdict.
  • assumptions on every result.

Engineering Method

The Python class FireWaterCoverageModel uses open, published relations only.

  • Demand is the application rate times the protected area, plus the sum of dedicated object surfaces times their rates, times a simultaneous-release factor.
  • Nozzle discharge follows the orifice law $Q = K\sqrt{p}$ with $K$ in $(\mathrm{l/min})/\sqrt{\mathrm{bar}}$.
  • The nozzle count is the larger of the flow criterion (total discharge must meet the density) and the coverage criterion (area divided by the square of the maximum permitted spacing). Sizing on flow alone is the classic error: it gives too few, too widely spaced nozzles that meet the average density but leave dry patches.
  • Monitor droplets are drifted downwind by $\Delta x = h,u_{\mathrm{wind}}/v_t$, and the coverage loss is taken as linear in the drift over the characteristic target dimension.
  • Shadowing is a flag, not a model: a monitor cannot wet a surface it cannot see.

Python Usage Pattern

from firewater_deluge_design import FireWaterCoverageModel

model = FireWaterCoverageModel()

demand = model.demand(
    protected_area_m2=420.0,
    area_rate_lpm_per_m2=10.0,
    objects=(("20-VA-001", model.horizontal_vessel_surface_m2(3.0, 9.0), 10.0),),
    duration_min=30.0,
)
print(demand.total_demand_m3_per_h, demand.object_to_area_ratio)

layout = model.deluge_layout(
    protected_area_m2=420.0,
    required_density_lpm_per_m2=10.0,
    nozzle_k_lpm_per_sqrt_bar=57.0,
    nozzle_min_pressure_barg=2.0,
    max_spacing_m=3.7,
    operating_pressure_barg=3.5,
)
print(layout.nozzle_count, layout.governing_criterion, layout.grid_spacing_m)

monitors = model.monitor_screening(
    target_area_m2=420.0,
    required_density_lpm_per_m2=10.0,
    monitor_count=2,
    monitor_flow_lpm=2400.0,
    wind_speed_m_s=12.0,
    fall_height_m=12.0,
)
print(monitors.effective_density_lpm_per_m2, monitors.verdict)

The Four Questions, in Order

Answer them in this order. Reversing the order is the usual way these studies go wrong.

  1. What fire are we protecting against? A pressurised gas jet fire is not extinguished by water; the creditable function is cooling of exposed surfaces and control of escalation. A liquid pool fire can be controlled and extinguished by foam-water. An area labelled "gas" that holds lube and seal-oil tanks still has the pool-fire case live, a common omission. Take the fire from neqsim-consequence-analysis.
  2. Can the inventory be removed faster than the target fails? NORSOK S-001 makes blowdown the primary means of protecting a pressurised inventory. Run DepressurizationSimulator against PfpDemandCalculator.bareSteelTimeToCriticalS(). Thick-walled vessels frequently outlive the blowdown; thin support steel frequently does not, and that asymmetry decides where protection goes.
  3. What does the requirement actually demand? See the application-rate table and the traps below, and read the edition of the standard carefully.
  4. Only then: size the water, lay out the nozzles, and test the hydraulics.

Java API (neqsim.process.safety.firewater)

Java 8, all classes Serializable, all emit schema-versioned JSON via toJson(). Do not use FireProtectionDesign.firewaterDemand(...) (neqsim.process.mechanicaldesign.designstandards) for these questions: it returns one lumped m³/h figure and carries no coverage, nozzle, monitor or hydraulic-feasibility logic.

// 1. demand: blanket area coverage
FireWaterDemandCalculator area = new FireWaterDemandCalculator(510.0)
    .setAreaRate(FireWaterDemandCalculator.NORSOK_PROCESS_AREA_LPM_M2)  // 10 (l/min)/m2
    .setDurationMin(30.0)
    .setFoamConcentratePercent(1.0);
double lpm = area.totalDemandLpm();          // 5100
double m3h = area.totalDemandM3PerHour();    // 306
double vol = area.waterVolumeM3();           // 153

// 1b. demand: dedicated object protection instead
FireWaterDemandCalculator obj = new FireWaterDemandCalculator(0.0).setAreaRate(0.0);
obj.addHorizontalVessel("HA-13-0001", 0.72, 6.606, 10.0);   // shell OD, length, rate
double saving = 1.0 - obj.totalDemandLpm() / area.totalDemandLpm();

// 2. nozzle net, governed by the LARGER of the flow and spacing criteria
DelugeNozzleLayout.Nozzle hv26 = new DelugeNozzleLayout.Nozzle(
    "HV26", 42.9, 3.5, 3.0, 1.0);   // K [(l/min)/sqrt(bar)], p_min, max spacing, clear depth
DelugeNozzleLayout net = new DelugeNozzleLayout(510.0, 10.0, hv26)
    .setOperatingPressureBarg(5.0)
    .setCoverageEfficiency(1.0);    // < 1.0 tightens the grid for wind or shadowing
net.nozzleCountFromFlow();          // 54
net.nozzleCountFromCoverage();      // 57
net.nozzleCount();                  // 57, coverage governs
net.governingCriterion();           // "coverage"
net.gridSpacingM();                 // 2.99
net.isClearanceAdequate(0.8);       // false for an HV nozzle needing 1.0 m

// 3. monitors as an alternative
FireMonitorCoverage mon = new FireMonitorCoverage(510.0, 10.0)
    .setMonitors(4, 4000.0).setGeometry(40.0, 120.0)
    .setWind(15.0, 12.0)                       // wind speed, droplet fall height
    .setCharacteristicTargetDimensionM(25.5)
    .setLineOfSightObstructed(true);
mon.nominalDensityLpmPerM2();     // 31.4, passes in still air
mon.effectiveDensityLpmPerM2();   //  3.7, fails once the pattern drifts
mon.verdict();

// 4. can the existing system take it?
FireWaterCoverageAssessment gate = new FireWaterCoverageAssessment(510.0, 5100.0)
    .setCoveredAreaM2(0.0)
    .setSpareSupplyFlowLpm(16748.0)
    .setWorstCasePressureMarginBar(-0.9, "XX-71-0010B")   // worst DESIGN case, not normal case
    .setRequiredPressureMarginBar(0.5);
gate.verdict();          // SUPPLY_PRESSURE_DEFICIT
gate.recommendation();

// 5. active vs passive: what the framework does and does not allow
ActiveFireProtectionScreening s = new ActiveFireProtectionScreening(
        ActiveFireProtectionScreening.FireScenario.GAS_JET_FIRE)
    .setIncidentHeatFluxKWPerM2(200.0)
    .setBlowdownTimeToTargetS(359.0)
    .setBareSteelTimeToCriticalS(98.7)
    .setLoadBearingStructureExposed(true)
    .setNozzlesCanBePlacedAboveEquipment(false);
s.isWaterExtinguishing();             // false for a gas jet fire
s.isPassiveSubstitutionPermitted();   // false, always
s.isBlowdownFasterThanFailure();      // false here: thin steel loses the race
s.rankedMeasures();

Validated by FireWaterDesignTest (orifice law, both nozzle-count criteria, wind-drift model, verdict routing, JSON contract).

Application Rates

BasisApplies toRate
NORSOK S-001 Ed.4 §20.4.4 (Ed.5/6 §21)Process areas and equipment surfaces10 (l/min)/m²
NORSOK S-001 Ed.4 §20.4.4Wellhead, riser balconies, FPSO turret manifolds20 (l/min)/m²
NORSOK S-001 Ed.4 §20.4.4Water at design pressure after a confirmed fire signal≤ 30 s
NFPA 15Exposure protection of vessel shells10.2 (l/min)/m² (0.25 gpm/ft²)
Common project practiceEnclosed utility/machinery rooms, sprinkler6 (l/min)/m²

NORSOK S-001 does not state a fire-water duration. Take it from the project basis (30 min is the usual foam figure and is often carried across) and say where it came from.

Traps

  • Clause renumbering. NORSOK S-001 Ed.4 (2008): clause 20 = fire fighting systems, clause 21 = escape and evacuation. Ed.5 (2018) / Ed.6 (2020): clause 20 = passive fire protection, clause 21 = active fire protection. A finding citing "21.4.4" cites Ed.5/6. Check which edition the installation's own basis invokes before agreeing that a requirement applies.
  • Fire water never substitutes for passive fire protection. The Norwegian Facilities Regulations §29 forbid crediting the cooling effect of fire-fighting equipment when designing PFP, and §11 forbids crediting fire water when establishing design accidental loads (NORSOK S-001 Ed.4 §19.4.2/§19.4.3 say the same). The converse is also not permitted: PFP on one item does not discharge an area fire-water requirement.
  • Flow alone under-counts nozzles. A high-K nozzle meets the density with few nozzles spaced far apart, leaving dry patches. Take the larger of the flow and spacing counts.
  • Take the pressure margin from the worst design case, not the normal case. Mature fire-water systems are commonly pressure-limited, and the margin can be negative in a two-pump case while looking healthy in the three-pump case.
  • "Nozzles cannot go above the equipment" is not a reason to have no coverage. Use side-mounted nozzles or a dedicated spray ring. A project basis that excludes coverage over an item is a deviation and must be recorded and accepted as one.
  • Monitors are not equivalent to a fixed system. NORSOK S-001 Ed.4 permits monitor substitution only for the turret, and §20.4.7 binds monitors to the same density and functionality. On an open deck, wind drift and shadowing usually settle the question.

Validation Checklist

  • The application rate comes from the governing project standard, not from the default: 10 (l/min)/m² for process areas, 20 for wellheads and riser balconies.
  • Object surfaces are exposed wetted surfaces, not plan footprints.
  • The nozzle count reports governing_criterion; a flow-governed count has been cross-checked against the maximum permitted spacing.
  • pressure_adequate and density_met are both true, or the deficit is stated.
  • The duration used for water_volume_m3 is traceable to the project basis.
  • Monitor results are reported with the wind speed they assume.
  • The result is presented as a screening input to a hydraulic network calculation.

Common Mistakes

SymptomCauseFix
Nozzle count too low, dry patches in the areaSized on the flow criterion aloneTake the larger of the flow and coverage counts; check governing_criterion
Higher-K nozzles do not reduce the countThe layout is spacing-governedReduce the protected area or accept the grid; K cannot fix overlap
Demand far below expectationObject surfaces entered as footprintsUse exposed wetted surface, e.g. horizontal_vessel_surface_m2()
Monitor concept looks adequateScreened at zero windRe-run at the design wind speed and report verdict
Passive fire protection reduced because deluge was addedSubstitution applied in the wrong directionKeep passive protection; the rule runs one way only

Interpretation

  • object_to_area_ratio well below 1 means selective protection of the hydrocarbon-bearing items is far cheaper in water than blanket coverage. Whether it is permitted is a requirement question, not a hydraulic one — the area-coverage requirement in NORSOK S-001 and ISO 13702 is written against the area.
  • governing_criterion == "coverage" means the nozzle net is set by spray overlap, so a higher-K nozzle will not reduce the count.
  • verdict == "marginal_wind_limited" means the monitor concept only works in still air.

Limitations

  • Screening only. It performs no network hydraulics, no trajectory modelling and no CFD.
  • It does not decide compliance. Application rates, durations and the acceptability of selective protection come from the governing project standard and the accepted deviations.
  • It gives no credit for, and takes no credit from, passive fire protection. Fire water and passive fire protection are complementary barriers, and the regulatory rule runs one way: passive protection may not be reduced because active protection exists.

Related Skills

  • neqsim-consequence-analysis — fire type, heat flux and flame reach (upstream).
  • neqsim-depressurization-mdmt — blowdown time, the primary barrier (upstream).
  • neqsim-technical-document-reading — area geometry, nozzle K-factors, hydraulic report (upstream).
  • neqsim-process-safety — barrier register, performance standards, LOPA (downstream).
  • neqsim-standards-lookup — clause traceability into results.json (downstream).
  • neqsim-jet-fire-radiation-screening and neqsim-relief-load-screening — the fire side.
  • neqsim-depressurization-screening — inventory removal, the primary barrier for a pressurised system.
  • neqsim-safety-function-coverage-screening — whether the protective functions exist at all.
  • neqsim-weather-data — live or forecast wind_speed_m_s for the fire-monitor wind-drift screening, instead of an assumed value.

References

  • NORSOK S-001, Technical Safety — fire-water application rates and area coverage.
  • ISO 13702, Control and Mitigation of Fires and Explosions on Offshore Production Installations.
  • NFPA 15, Standard for Water Spray Fixed Systems for Fire Protection — nozzle spacing and density.
  • API RP 2030, Application of Fixed Water Spray Systems for Fire Protection in the Petroleum and Petrochemical Industries.
  • NeqSim repository: https://github.com/equinor/neqsim

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.