Fire and Flood Get Screened Twice

April: the site clears the fire and flood screen. October: a conference call about foam concentrate inventory, hazmat team travel time, and whether the responding department has ever worked a high-piled storage fire. Nobody at the economic development organization can account for why the file reopened.

It reopened because two evaluations were always scheduled. The city prepared for one of them.

The first belongs to the site selector. It runs early, off public record, often before anyone locally knows a project exists. Four artifacts: the parcel's ISO Public Protection Classification, a hydrant flow test stated in gallons per minute at 20 psi residual pressure, station drive distance and mutual aid coverage, the FEMA flood map panel and its effective date. Pass or fail. All four can sit in a folder before the RFI arrives, sourced from officials who already hold the underlying data.

The second belongs to the prospect's insurer or corporate risk engineering team, and it runs against a specific building holding specific material. Sprinkler design basis against what gets stored and how high. Available fire flow against required density and duration. Hazmat and foam capability against process chemistry. Flood exposure against the value of inventory and process equipment sitting on the slab. The output is a premium. Sometimes the output is harder than a premium.

The two run months apart and they ask different questions. Clearing the first says almost nothing about the second.

Whether this dimension governs your pipeline

Start with the archetype. It sorts faster than any sector list. Class 5 community, career department, 1,500 gpm available at the hydrant serving the marketed parcel. One water main. No Community Rating System participation. The parcel maps Zone X, and nobody has checked the access road.

That city is not 60 percent ready. It is fully competitive for ordinary-commodity manufacturing, and it cannot serve high-piled plastics storage on municipal supply alone. That tenant funds an on-site tank and fire pump or it sites elsewhere. The delta in required fire flow runs better than two to one, and none of it is a function of the building. Partial readiness earns no partial credit here. It creates lanes. An EDO that knows which side of the line its parcel sits on targets accordingly. An EDO that markets the number without the boundary loses to a site carrying full proof.

The dimension is deal-critical when insurable value concentrates in stored or processed material rather than in the structure:

• High-cube distribution with real commodity value
• Food and beverage, where combustible packaging stacks above ammonia refrigeration
• Coatings, adhesives, resins, chemical processing
• Aerosols and ignitable liquids
• Battery manufacturing and grid-scale storage
• Anything warehousing plastics at height

It is secondary for light assembly, machine shops, precision fabrication, and most contract manufacturing under 100,000 square feet with thin inventory. Those prospects ask the screening questions. They rarely fail them.

The city-side variables that decide it are few and knowable. Whether the marketed parcel sits inside the credited protection area or in a split classification that drops it below your headline number. Main diameter and pressure zone serving the site, storage volume, looped or dead-ended. Career versus volunteer staffing, and actual drive distance from the station that responds. Adopted fire code edition. Flood zone of the parcel. Flood zone of the access road. The last two are one question in most site profiles and two questions in underwriting.

Why the commodity decides the water demand

Two FM data sheets carry the arithmetic.

FM's Data Sheet 8-1, revised October 2025, classifies a stored product together with its internal packaging, external packaging, and pallet. Packaging frequently presents the worse fire challenge. Data Sheet 8-9 then derives protection from commodity class, storage arrangement, storage and ceiling height, sprinkler type and pressure, the number of sprinklers assumed to operate, hose stream allowance, and duration. Hose stream allowance is water reserved for firefighters on hand lines, demanded on top of what the sprinkler system draws.

One illustration off the DS 8-9 tables. Thirty-foot ceiling, wet system, quick-response pendent K25.2 extended-coverage sprinklers, two tenants.

Same building. Instantaneous demand up by a factor of 2.35, water volume up by a factor of 4.7, determined entirely by what the tenant decided to store. These are table-derived illustrations, not designs. They exclude elevation and friction losses, safety margin, pump and tank configuration, actual rack geometry and flue spaces, and whatever the insurer's own engineering concludes.

The caveats don't touch the point. Your fire flow figure is an input to a hydraulic calculation that cannot be run until someone names the commodity and the building. Structurally it is the wastewater screen again, where a sewer line proves nothing until the receiving plant's chemistry limits meet the prospect's actual effluent. At Rome, Georgia's Enterprise Corner, a funded 16-inch main proves water is being physically extended. It proves nothing about pressure, flow, duration, or sprinkler suitability for a tenant nobody has met.

Stage one, and the five documents it takes

Area Development's Q1 2026 consultant survey reports the profession reweighting away from cost optimization toward operational certainty and risk mitigation: site readiness, due diligence, permitting, infrastructure, competent local response. The same survey does not score property insurance cost, protection class, fire flow, or flood exposure as separate factors. The omission is not evidence that risk stopped mattering. Risk documentation is being absorbed into the readiness screen instead of tracked as its own category. Which means nobody calls to tell you which missing document sank the site.

The ISO report and the department file. ISO grades communities on a nominal 100-point Fire Suppression Rating Schedule: 50 points for fire department capability, 40 for water supply, 10 for emergency communications, plus a 5.5-point community risk reduction bonus. Class 1 begins at 90 points. Each class spans roughly ten points, down to Class 10.

Three things most portfolios get wrong.

1. It is address-specific. ISO's commercial reports identify protection area, responding station, station drive distance, and water supply type for a given location. A parcel at the edge of your service area can land in the Class 10 or 10W half of a split classification while your marketing sheet advertises Class 3. You can run the selector's test yourself in about ten minutes, before the report comes back. The credited-location thresholds are distances: within five road miles of a recognized responding station, within 1,000 feet of a creditable hydrant. Measure both to the marketed parcel. Not to the front gate of the industrial park.
2. It measures first-alarm structure fire suppression only. Not multiple alarms, not simultaneous incidents, not the department's full capability. That boundary is the reason stage two exists.
3. For large industrial risks it may not govern. Properties above 3,500 gpm ISO Needed Fire Flow can be evaluated separately and assigned an individual classification. If your target sector builds in that range, the community class is context, not conclusion.

File the report alongside CAD-derived response times to the parcel, an apparatus and staffing summary, and executed automatic and mutual aid agreements. Supply the yardstick with the data. NFPA 1710 sets deployment benchmarks for career departments, commonly applied as 240 seconds travel time for the first-arriving engine and 480 seconds for the full first-alarm assignment, both measured at the 90th percentile. NFPA 1720 governs volunteer and combination departments and sets staffing and response against demand-zone population density rather than a single travel figure. Confirm the current edition with your chief. A city that publishes its own percentile against the applicable benchmark, misses included, is doing something almost no competing site does.

The flow test. NFPA 291 is current in its 2025 edition, superseding 2022. A report a third party can actually use carries more than one number: test and flow hydrant identifiers, main size and whether looped or dead-ended, static pressure before flow, residual pressure during stabilized flow, outlet size and discharge coefficient, pitot or metered reading, measured GPM, computed flow available at 20 psi residual, date and time, conducting party, and operating conditions bearing on the result.

Keep the observed readings. The sprinkler designer checks the arithmetic, and NFPA 13 separately requires gauge location and elevation relative to the riser as part of design water supply information. A hydrant number and a bare "3,200 GPM at 20 psi" is a marketing line.

The perishability rule is the part cities get wrong. NFPA 13 requires waterflow data used for sprinkler design to be no more than 12 months old at working plan submission unless the AHJ accepts otherwise. That is a project-stage rule. It does not put an annual expiration on your inventory test. The defensible standing posture is dated screening evidence, valid while the network is materially unchanged, carrying a written refresh trigger: a new major user on the pressure zone, a main extension, a tank or pump change, a prolonged valve outage. Put the trigger sentence in the document. A test dated eighteen months back with a no-material-change statement from the water director reads as competence. An undated test reads as neglect.

Testing authority varies. Some utilities run their own crews; others let qualified fire protection contractors test with utility staff witnessing. Either way the controlling water authority issues or countersigns the record, because the operation moves its distribution system.

The sprinkler design basis, for every existing building you market. Greenfield sprinkler design is irreducibly prospect-specific. Spec buildings and existing product are not. Those buildings already carry a design basis, and it is a document: sprinkler type and K-factor; whether the system is ESFR, control-mode specific application, or a control-mode density-and-area design; design density and remote area; ceiling clear height; in-rack provisions if any; fire pump presence and rated capacity; on-site storage tank volume; hydraulic nameplate data at the riser.

Hold it because it tells you which commodity classes the building can already host. This is the lanes argument at the asset level. A high-clear building on a modern ESFR design serves one tenant set. The same footprint on a 1980s density-and-area system serves a smaller one, and nobody quotes that difference as a renovation line item from memory. Sources: the building owner's insurance file, the last sprinkler contractor, the fire code official's plan review file. Photograph the nameplate. Attach the as-builts and the calculation summary.

You can document the as-built design basis. Whether it suffices for a given tenant's commodity is the insurer's call.

The flood file. FEMA's September 2025 flood loads guide specifies what a defensible citation contains: FIRM panel number with revision suffix, effective date, flood zone, base flood elevation where published, vertical datum for every elevation, Flood Insurance Study number and effective date. Add all effective Letters of Map Change and a parcel-specific determination signed by your floodplain administrator.

Two map-change instruments get conflated routinely. A Letter of Map Amendment covers land or a structure naturally above the base flood elevation that appears inside the mapped hazard area. No FEMA review fee, but the applicant pays a surveyor or engineer to certify elevations. A Letter of Map Revision actually revises the effective map, requires hydraulic and hydrologic engineering, and carries FEMA review fees. FEMA's stated targets are 60 days after a complete amendment request and 90 days after a complete revision request. "Complete" is carrying the weight in that sentence.

Community Rating System status, if you have it, runs from Class 10 at no discount to Class 1 at a 45 percent NFIP premium discount, five points per class. Roughly 1,520 communities participate. Do not cite FEMA's eligible-communities list. It sweeps in Class 10 jurisdictions receiving nothing.

Two 2026 conditions to track. NFIP authorization is scheduled to lapse September 30, 2026 absent extension. And under Risk Rating 2.0, the FIRM zone is no longer the principal NFIP rating category, though the effective map still controls floodplain regulation and mandatory purchase. Neither changes what you document.

The adopted code. Fire and building code edition, local amendments, named fire code official, standard review process. Cheapest document in the portfolio, and vintage varies more than cities assume. PNNL's permitting guidance notes that energy storage provisions entered the International Fire Code in 2018 and were updated in 2021 while some jurisdictions still operate under 2015 or earlier editions. NFPA 855 is current in its 2026 edition: broadened hazard mitigation analysis, expanded large-scale fire test guidance, strengthened emergency response planning. A developer's engineer checks your adopted edition and amendments early. Keeping the answer current costs nothing.

Stage two, where the ISO class stops governing

Calling stage two everything ISO ignores overstates it. The schedule credits apparatus deployment, ladder service, staffing, training, pre-incident planning, automatic aid, and credited alternative water supply operations. Limited credit versus site-specific adequacy is the cleaner framing.

A risk engineer working against a named facility wants:

• The hazmat team's operational level, per-shift staffing, and travel time to the parcel, plus whether that team can handle the specific chemicals on the process flow diagram
• Foam agent type and quantity on hand, proportioning equipment, application rate, material compatibility, mutual aid resupply, runoff containment
• Whether a current pre-incident plan exists for this occupancy covering shutoffs, sprinkler control valves, drainage, and responder access
• Aerial reach measured from where apparatus can actually set up after setbacks, parked trailers, overhead lines, and roof geometry
• Where hydrant coverage is thin, a measured sustained tanker shuttle flow to the parcel rather than a theoretical one
• Whether the department has ever worked a comparable occupancy: combustible dust, high-piled storage, ammonia refrigeration, large-format lithium-ion

Flood diverges here harder than fire, and more expensively. FEMA's map governs floodplain regulation and mandatory purchase. It does not govern the insurer's price or its appetite. Commercial property carriers run their own catastrophe models against their own asset data, which the Urban Land Institute's April 2025 underwriting guide documents across more than fifty hazard-specific underwriting inputs. A parcel sitting cleanly outside the Special Flood Hazard Area clears the screening pass and still comes back flagged in October, because the engineer looked past the 1-percent-annual-chance event to lower-probability exposure against concentrated inventory value. The 0.2-percent data sits in the Flood Insurance Study you are already pulling. Know the parcel's relationship to it before the prospect's engineer computes it. Then run the same overlay on the access road, the utility corridors, and any pump station or substation the project depends on. A pad above the base flood elevation does not establish that trucks reach it during an event. Business interruption is where that gets priced.

None of this contradicts your class. ISO's schedule confines the classification to first-alarm structure fire suppression. Nor can any of it be pre-authorized as a commitment, because the questions take their shape from a building that does not exist yet. The counterparty mapping piece sorted information into public, NDA-ready, and project-specific classes. Insurer conclusions live irreducibly in the third.

Response capability does get fixed in writing sometimes, just not by an insurer. The Massachusetts siting board's June 2026 Oakham decision on a 180-MW battery project directs the developer to design to NFPA 855-2026, deliver its final hazard mitigation assessment and emergency operations plan at least 60 days before construction, deliver the final emergency response plan at least 60 days before commercial operation, coordinate with Oakham and mutual aid departments, and provide responder training and drill protocols. Regulatory condition, not insurance. Same shape.

The premium question, and its limits

I went looking for a defensible dollar figure to set beside a typical incentive package. It is not in the public record.

What exists measures the wrong thing. Institutional property data put industrial insurance expense at $0.055 per square foot per quarter in Q2 2023, up 88 percent from Q4 2019. Roughly $0.22 annualized. That tracks the building. It captures none of a manufacturer's inventory, process equipment, business interruption, product, or pollution coverages, which is exactly where the money sits for the prospect types this dimension governs.

No credible national differential exists by protection class, sprinkler status, or catastrophe exposure. Price depends on total insured value, inventory and equipment concentration, business interruption values, occupancy and commodity, construction, private fire protection, catastrophe model output, limits, deductibles, loss history, and the insurer's own engineering findings. FM reports more than ten times the average gross loss per $100 of value in unsprinklered facilities within its managed claims data. That is a loss statistic. It should not be converted into a premium ratio.

Price is not the only outcome, either. Below some threshold of protection, response capability, or catastrophe exposure, a site stops being expensive and becomes difficult to place: the program moves out of the standard market into surplus lines, or it comes back with percentage deductibles and flood sublimits a CFO will not sign. That is appetite, not pricing, and it kills projects more quietly than a premium quote does.

The mechanism is documentable. The magnitude isn't. Protection quality, response capability, and catastrophe exposure enter the prospect's property program as recurring annual cost. That cost is permanent. Your incentive usually isn't.

Four rungs and three emails

What a city can pre-authorize is the process. Four rungs:

• Observed condition. Dated flow test, address-specific ISO report, sprinkler design basis for existing product, FIRM citation, adopted code edition.
• Indicative capability. Apparatus and staffing summary, CAD response times against the applicable benchmark, executed aid agreements, credited shuttle capability.
• Defined test protocol. What the city will test once commodity, footprint, storage height, and sprinkler concept are known. On what timeline, at whose cost, signed by whom.
• Test completed. Project-stage flow test meeting the 12-month rule, results in the sprinkler designer's hands.

Stopping at rung one and calling it readiness is the standard failure. Reaching rung three before a prospect is named is the opportunity, and it is a writing exercise, not a capital project.

To the water director. Issue a signed PDF hydrant flow report for the marketed parcel under the current NFPA 291 configuration: test and flow hydrant identifiers with a sketch or map, main size and whether looped or dead-ended, date and time, static and residual pressure, outlet size and coefficient, pitot or metered reading, observed GPM, projected flow at 20 psi residual, operating conditions during the test, testing personnel. State in the document which system changes would require a new test. A sprinkler engineer has to re-derive the curve from the observed readings, so a single summary number cannot be used for design and will trigger a retest at the worst point in the schedule.

To the fire chief. Provide the current address-specific ISO PPC report for the parcel, including any split classification, responding station, and drive distance. Provide station, apparatus, and staffing summary; CAD response time data for the site; executed automatic and mutual aid agreements; credited water shuttle or alternative supply capability; and current commercial pre-incident planning practice. Name the official, with role-based backup, authorized to discuss project-specific hazmat, foam, aerial access, and pre-planning requirements under NDA. The address matters because ISO classifies by location, and an underwriter comparing a city-authored summary against ISO's own record treats the discrepancy as the finding.

To the floodplain administrator. Provide an effective FIRMette and a parcel-specific determination stating full panel number with suffix, effective date, zone, base flood elevation where applicable, vertical datum, relevant Flood Insurance Study material including 0.2-percent-annual-chance data, all effective Letters of Map Change, and current CRS class. State whether the parcel, the access road, or the planned utility corridors would require further elevation certification or hydraulic work, and identify who bears that cost. The prospect's insurer models beyond the regulatory floodplain, so a determination limited to the building pad answers half the question that gets priced.

Three emails produce a file that answers stage one completely and sets up stage two honestly. The cities that lose this dimension are rarely the cities with bad hydrants.