Water Damage Restoration in Middlesex County, NJ
Perth Amboy and Woodbridge flood on a tide table. Spotswood and Old Bridge flood on a rain gauge. Our own crews, either way — extraction, drying, mold remediation and the rebuild.
Middlesex is two counties as far as water is concerned, and the county's own claim record says so. Of the 5,251 National Flood Insurance Program claims FEMA holds under this county's code, 2,378 name accumulation of rainfall or snowmelt as a cause, 1,848 name a stream, river or lake overflowing, and 762 name tidal water. Those are three different ways of getting wet, and the third one runs on a clock rather than on the weather.
On the east, the Arthur Kill, the lower Raritan and the Raritan Bay push water back up the tidal creeks into Perth Amboy, Woodbridge, Sayreville, South Amboy and South River. On the west, rain falls on sand and reaches Manalapan Brook, Lawrence Brook and Deep Run fast, crests, and is gone inside a day. What that means for a building is different in each case, and it is the part of this job worth explaining before anything else.
This page does not repeat the general parts. What to photograph before anything moves, what a flood policy does and does not reach inside a basement, and when the rebuild needs a permit are all written out on our water damage restoration page for New Jersey. What follows is Middlesex.
What we do
- Water extraction and structural drying — our own crews, not a dispatch network
- Mold remediation — remediation of what the water left behind
- Sewage and flood cleanup — water that came up through a sewer or across the ground
- Rebuild after the damage — the same company puts the building back
- All 21 New Jersey counties — the shop and the crews are in Garfield, Bergen County
- Paid by insurance or by you — we document the loss and the work; we do not handle the claim
A tide flood and a brook flood are not the same job
A tidal flood is scheduled. It arrives on the high water, it leaves on the low, and if the wind is still on shore it comes back roughly twelve and a half hours later. That has one practical consequence people learn the hard way: extraction started in the middle of a tidal event gets undone by the next one. The work that actually pays off between tides is getting contents up and out and getting power decisions made, and the extraction starts when the water has stopped coming back.
A brook flood is the opposite. It arrives fast, it peaks once, and it is usually falling before anybody has finished making phone calls. There is no second wave to wait out, which means the 24 to 48 hour drying window the EPA writes about starts immediately, not tomorrow.
The water is different too. Tidal water off the Arthur Kill and the Raritan Bay is brackish and it carries silt; it leaves salt in materials that then hold moisture and stay conductive long after the surface reads dry. Brook water is fresh but it has run across ground first. Which of those a building took decides what comes out and what can be dried in place — and that determination is made by looking at the water and the materials, not by reading the weather report, which is a point the standard itself had to publish a correction about.
The Arthur Kill gauge is written at county scale, which is unusual
Most river gauges publish flood text about one bridge and one road. The NOAA gauge on the Arthur Kill at Perth Amboy (PABN4) does something rarer: its major-flood statement is written about Middlesex County as a whole. Minor flooding is set at 7.2 feet above mean lower low water, moderate at 8.2, and major at 9.2 — and at 9.2 feet the published text says widespread roadway flooding begins in Middlesex County near the Arthur Kill, the lower Raritan River, the Raritan Bay and the tidal tributaries, with numerous roads impassable and some neighbourhoods isolated.
The more useful number is the one below it. At 7.8 feet — above the minor threshold of 7.2, but four-tenths of a foot below the moderate stage of 8.2 — the same gauge names the places where it starts: Woodbridge Creek in Woodbridge Township and Perth Amboy, especially around State Street; the South River in East Brunswick, South River and Sayreville; Victory Plaza in Sayreville where Route 9 meets Route 35; North Rosewell Avenue in South Amboy; and Route 35 in Laurence Harbor. If you live on that list, a forecast tide of 7.8 feet is your number, and it is four-tenths of a foot under the level at which anybody issues an advisory.
The gauge's record is 9.39 feet, and it has been reached twice — on 27 October 2018 and again on 23 December 2022. Neither was a hurricane. That is worth sitting with, because the storms people plan around are not the ones that have topped this gauge in the last decade.
Inland: sand, a fast crest, and a tide that still gets a vote
Manalapan Brook at Spotswood (SPTN4) reaches its action stage at 18.5 feet and floods at 19, where the published impact is that Mundy Avenue, which connects Spotswood and Old Bridge, is subject to flooding and closure. The same statement adds a sentence you will not find on any other inland gauge in this county: downstream tides can affect Lake Devoe levels and the flooding that follows. The tide reaches thirteen miles inland, and the brook backs up behind it.
The peak of record there is 22.22 feet, carrying 6,650 cubic feet per second, on 28 August 2011 — the largest of 69 annual peaks the USGS has measured at that site since 1957. Lawrence Brook at Westons Mills has a shorter book, 35 annual peaks from 1989, and the same two storms sit on top of it: 20.26 feet at 7,700 cubic feet per second in August 2011, then 19.81 feet at 6,430 in September 2021.
One gap is worth knowing about, because it changes what you can look up. The South River is named in the tide gauge's own flood text, but the USGS gauge on it at Old Bridge stopped recording in February 1988. Its 49 annual peaks run from 1940 to then, and the four highest — September 1944, May 1968, August 1971 and May 1984 — sit within fourteen hundredths of a foot of one another at 11.59 to 11.73 feet. Nothing in that record covers Floyd, Irene, Sandy or Ida. If somebody tells you what the river did in 2011, they are not reading it off that gauge, because that gauge was not running.
What the claims record shows, and who is not in it
Two things stand out in the county's file. The first is that 3,832 of the 5,251 claims carry a basement or enclosure code at all — about three in four, a far higher share than the state manages — and of those, 3,528 are on a building with a finished or unfinished basement or enclosure. Middlesex floods into rooms people were using.
The second is the policy count. There are 2,496 residential flood policies in force against 212,397 residential structures: 1.18 percent, the thinnest of the eight counties with a page here. Inside the mapped high-risk zone the picture is ordinary — 1,280 policies against 2,391 structures, 53.5 percent. Outside it, 1,216 policies cover 210,006 structures, which is 0.58 percent. The claims below are therefore a sample of a very small minority, and the number of Middlesex buildings that have actually taken water is much larger and is not recorded anywhere.
The Community Rating System column is the one homeowners misread most often. All 25 Middlesex municipalities take part in the flood insurance programme, but only three hold a Community Rating System class: Woodbridge at class 5, which earns a 25 percent discount on policies inside the mapped zone, Perth Amboy at class 9 for 5 percent, and South River at class 10, which earns nothing at all. That is a measure of what the town's floodplain office does with its paperwork. It is not a measure of how wet your street is.
| What was counted | Middlesex County | Basis |
|---|---|---|
| NFIP claims on record, 1978 to 2026 | 5,251 | OpenFEMA NfipClaims v3, county code 34023 |
| Of those, closed with a payment | 4,384 | Building or contents payment above zero |
| Paid on those claims | About $164.6 million | Summed off the claim rows: building, contents, compliance |
| Claims naming accumulated rainfall or snowmelt | 2,378 | FEMA cause-of-damage code 4 |
| Claims naming stream, river or lake overflow | 1,848 | Cause code 2 |
| Claims naming tidal water overflow | 762 | Cause code 1 |
| Claims carrying any basement or enclosure code | 3,832 of 5,251 | The field is blank on the other 1,419 |
| Of those, a finished basement or enclosure | 1,404 | Basement code 1 |
| Of those, an unfinished basement or enclosure | 2,124 | Basement code 2 |
| Worst single years by claim count | 2011 (1,114), 2012 (747), 2021 (712) | Year of loss |
| Named events | Irene 1,071, Sandy 742, Ida 616, March 2010 nor'easter 465 | FEMA's own event names |
| Residential flood policies in force | 2,496 of 212,397 structures (1.18%) | FEMA penetration rates file, as of 3 August 2026 |
| Inside the mapped high-risk zone | 1,280 of 2,391 (53.5%) | Same file |
| The same, outside it | 1,216 of 210,006 (0.58%) | Arithmetic on the same row |
| Municipalities in the flood insurance programme | 25 of 25 | Community Status Book, refreshed 2026-08-25 |
| Holding a Community Rating System class | 3 — Woodbridge 5, Perth Amboy 9, South River 10 | Class 10 earns no discount |
| Housing units built before 1980 | 182,573 of 317,979 (57.4%) | ACS 2020-2024 five-year table B25034, the 007 to 011 rows |
How to read the cause rows: FEMA's cause-of-damage field allows more than one code on a claim, so these counts are each a count of claims naming that cause and they are not meant to add up to 5,251. How to read the claims total: it counts insured buildings only. With 1.18 percent of this county's residential structures insured for flood, every figure above describes the insured minority, and the uninsured majority is not in any file.
What the first day looks like on each side of this county
On the tidal side, the first hour is triage rather than extraction, for the reason above: the water is coming back. What we do in it is get a record made and get the salvageable contents out of the wet rooms, and then extract once the tide has turned for good. After brackish water, the drying target is the same but the cleaning is not — salt and silt stay in porous material after the water leaves, and a wall cavity that has held bay water dries slower and reads differently than one that held rain.
On the inland side the clock is the only thing that matters, because the crest has already passed by the time most people call. Extraction, then everything that cannot be dried in place comes out, then the structure is dried until the readings stop moving rather than until the surface feels right, and somebody comes back to check that they have.
Then the rebuild, and it is the same company that did the drying. Because so much of this county's damage sits in finished basement space, the rebuild is usually the bigger half of a Middlesex job: what was lost was a room somebody was using, and putting that back is carpentry rather than extraction. You can ask us for a written estimate before any of it, and the documents you should end up holding are listed on the statewide page.
- Tidal event: record it, lift what can be saved, extract after the water stops returning.
- Brook flood: extract immediately — the crest has passed and the drying window is already open.
- Brackish water leaves salt and silt behind, so cleaning and the drying check both take longer.
Common questions
My street floods on a big tide and drains by itself a few hours later. Is there anything to do?
Yes, and the reason is what the water leaves rather than how long it stayed. Brackish water off the Arthur Kill and the Raritan Bay carries salt and silt into porous material, and both stay after the water has gone: salt holds moisture, silt holds everything else. A cavity that has been wetted by bay water three times in a winter is not in the same condition as one that has been wetted once by a burst pipe. What is worth doing is getting the wet material looked at before it is closed up again, which is a much smaller job than the one that follows if it is not.
Woodbridge is a Class 5 community. Does that mean my house is safer than one in the next town?
No. A Community Rating System class is a discount on flood insurance premiums that a municipality earns for its own floodplain management work — mapping, ordinances, outreach, record-keeping. Woodbridge's class 5 is worth 25 percent off policies inside the mapped high-risk zone, Perth Amboy's class 9 is worth 5 percent, and South River's class 10 is worth nothing. It says something real about the town hall. It says nothing about the elevation of your floor, which is on the FEMA flood map rather than in the rating.
Sources
- FEMA, OpenFEMA NFIP Redacted Claims (v3), Middlesex County rows (county code 34023) — retrieved
- FEMA, NFIP Residential Penetration Rates, Middlesex County row (as of 2026-08-03) — retrieved
- FEMA, NFIP Community Status Book, the 25 Middlesex County communities (refreshed 2026-08-25) — retrieved
- US Census Bureau, American Community Survey 2020-2024 5-year, table B25034, Middlesex County — retrieved
- NOAA National Water Prediction Service, Arthur Kill at Perth Amboy (PABN4) stages, impact statements and crests — retrieved
- NOAA National Water Prediction Service, Manalapan Brook at Spotswood (SPTN4) stages and impact statements — retrieved
- USGS, annual peak streamflow, Manalapan Brook at Spotswood (01405400), 69 annual peaks from 1957 — retrieved
- USGS, annual peak streamflow, Lawrence Brook at Westons Mills (01405030), 35 annual peaks from 1989 — retrieved
- USGS, annual peak streamflow, South River at Old Bridge (01405500), 49 annual peaks, 1940 to 1988 — retrieved
Water in the building?
Tell us what happened and what you can see. Pick “Water Damage” and we will come back to you about extraction, drying and what the rebuild involves.