Strip Cooling Troubleshooting: Why the Coil Cools Uneven and What to Check First

The strip leaves the runout table at the right average temperature, and the coil still fails flatness on the temper mill. When the cooling side of the line is the suspect, the investigation almost always follows the same path: the average temperature is fine, but the distribution of temperature across the width, along the length, or between passes is not. Strip cooling troubleshooting is really a search for where the water is not going where you think it is.

This guide maps the five failure families that show up in strip cooling plants, transverse spread, edge overcooling, low pressure, clogged nozzles, and nozzle wear, each to a symptom–cause–remedy table, then a field inspection procedure and a maintenance plan that keeps most of these faults from ever reaching a coil.

Why Cooling Uniformity Is a Metallurgical Requirement

Cooling sets the hot strip’s final microstructure: the rate at which austenite transforms determines grain size, phase fractions, and therefore yield strength, ductility and hardness. A strip that cools at one rate in the middle and another at the edges is literally two different materials across its width.

The heat to be removed is arithmetic. Sensible cooling follows Q = m·cp·ΔT: water’s specific heat is about 4.18 kJ/kg·K, so each litre rising 10 °C absorbs roughly 42 kJ. Evaporation is far more powerful. The latent heat of vaporization is about 2.26 MJ/kg, so a kilogram that fully evaporates removes more than fifty times the energy of the same kilogram warmed by 10 °C. That is why mist-assisted zones do so much work with so little water, and why a nozzle that makes coarse drops that bounce off a hot surface quietly kills the cooling rate of the whole zone.

A 20 °C spread across the width can mean a coil that reels flat but develops wavy edges once cut, because regions that cooled at different rates respond differently to the same downstream strain. Strip cooling troubleshooting starts from one principle: find the non-uniformity in the water, because the steel only does what the water tells it to do.

Read the Pattern Before You Touch a Nozzle

The fastest diagnostic step is free: read the defect pattern before anyone climbs onto the header.

A band repeating at exactly the nozzle pitch points at the spray row: wrong spacing, wrong spray angle, or a systematic overlap error.
One cold or hot lane at a fixed position points at one or two nozzles: a clogged or missing tip, a misaligned nozzle, or a worn orifice.
Edge-only differences point at edge effects: run-off, edge masking, or nozzles set to spray past the strip edge.
A shift that moves with line speed or pump speed points upstream: pump, filters, pressure control or water temperature.
A defect that appeared after a nozzle change or header repair points at what was touched: wrong nozzle, wrong orientation, or a header re-levelled badly.

These five patterns cover most strip cooling troubleshooting cases; the tables below map each family in full.

Transverse Temperature Spread Across the Strip Width

The classic fault is a strip cold in one zone and hot in another: a gradient from edge to edge, or repeating hot/cold banding.

Symptom	Likely cause	Quick check	Remedy
Steady cold band on one side of the strip	Nozzles on that side at the wrong angle or height; header misalignment	Measure header height and nozzle angle at both ends	Re-level the header; reset nozzle angles to design
Repeating hot/cold bands at nozzle pitch	Overlap too small or too large for the spray angle and height	Check nozzle pitch against pattern width at the strip surface	Adjust header height or spacing so adjacent fans overlap correctly
Gradient edge to edge with no banding	Supply imbalance: header fed from one end only, pressure decays along its length	Compare pressure at both header ends	Feed from both ends or enlarge the header
Hot lane that wanders from coil to coil	Cooling water temperature rising through the day; low tank level	Check supply temperature against the log	Fix water temperature control; top up or clean the recirculation circuit
Banding after a nozzle change	Mixed nozzle types on one row: tapered-edge and even-edge fans overlap differently	Walk the row and check angle and edge type on each body	Standardize the row to one nozzle type and angle

The overlap failure is simple arithmetic. Each flat fan spray nozzle lays down a density profile highest at the centre and falling toward the edges; where two fans meet, the local intensity is the sum of the two profiles. Too little overlap and the sum dips, a hot seam; too much and it peaks, a cold band. The same nozzle can give perfect uniformity at one header height and banding at another, which is why the first remedy is geometry, not parts.

Edge Overcooling and Edge Run-Off

Edges behave differently for two reasons. Heat escapes through the side face as well as the top and bottom faces, so the edge runs colder for the same water. And a pattern sized for the full strip width spills water past the edge, creating an over-cooled rim that drives wavy-edge defects.

Symptom	Likely cause	Quick check	Remedy
Wavy edges (edge wave) on the finished coil	Edge overcooling: more water per unit area at the edge than the middle	Compare measured edge vs centre temperature at each zone exit	Trim the outer nozzles so the pattern stops just inside the strip edge
Colder edges only in the finishing zone	Nozzles spraying past the strip edge: pattern wider than the strip	Measure the wetted width at the strip surface	Smaller spray angle, lower header, or edge shields
Edge cooling varies when strip width changes	Header set for the widest strip; narrower coils get over-wetted edges	Check the row setting for the last narrow product	Add edge-trimming nozzles, opened or closed by product width
Colder edges on one side only	Header skew: one side of the row closer to the strip	Measure both header ends to the strip	Re-level; re-check after every header repair
Edge wave appears only at high line speed	Run-off increases with speed; the edge film thickens and overcools	Re-run the speed/edge-temperature correlation from the log	Trim edge-zone flow with line speed, or add edge masking

The standard answer is deliberate edge management: keep the middle uniform and give the edge its own trim zone with valves, tuned to width and speed. If your line has no trim zone, set the last nozzle at each end so its pattern ends at the strip edge. An uncovered hot rim is easier to correct than a hard, over-cooled edge.

Low Header Pressure or Insufficient Flow

When the cooling section runs short of water, the strip exits hot and coiling temperature cannot be held. Low pressure is usually a system problem, not a nozzle problem: the nozzles faithfully convert whatever pressure they see into flow.

The governing relation is the square-root law: nozzle flow is proportional to the square root of the pressure drop across the orifice. Cut header pressure by 20% and flow drops roughly 11%; cut it by 30% and flow drops about 16%. Because cooling effect tracks water delivered per unit area, a small pressure sag quietly pushes the strip out of its window.

Symptom	Likely cause	Quick check	Remedy
Header pressure below design at full demand	Blocked filter or strainer; undersized supply line; pump wear	Read filter pressure drop; compare pump curve to duty	Clean or replace filters; check the pump impeller; open supply valves
Pressure sags as more zones open	Supply capacity is the limit: pump, header or line cannot feed all zones at once	Record pressure with one zone vs all zones open	Sequence zones instead of running all at full; upgrade pump or header feed
Pressure stable but strip runs hot	Water too warm, or nozzles worn and the pattern has widened	Check supply temperature; bucket-test a sample of nozzles	Cool the recirculation water; replace worn nozzles
One zone under-cools at normal header pressure	Zone isolation valve partly closed; internal header scale; air lock	Check the zone valve position and local pressure at that zone	Open the valve; vent the header; descale if fouling is confirmed
Gauge reads fine but flow is low	Wrong gauge location, or a blocked impulse line giving a false reading	Cross-check with a second gauge or a flow measurement	Clean or relocate the gauge; install a flowmeter on the main supply

The supply-side diagnosis follows a ladder: confirm the pump delivers rated flow, the filter is clean, and the header is fed from the correct end, then look at the nozzles. Each rung eliminates a whole class of causes, and each is cheaper than the next.

Clogged and Partially Blocked Nozzles

Clogging is the most common cause of a local strip cooling fault, and the most misdiagnosed: a partially blocked nozzle changes the pattern long before it stops spraying. Scale, rust, sand and biological growth all accumulate in the smallest cross-section of the system: the orifice and the internal vanes that shape the spray.

Symptom	Likely cause	Quick check	Remedy
One cold lane at a fixed position	A single nozzle partially clogged, spraying a distorted or reduced pattern	Visual check of that nozzle’s spray against its neighbours	Remove and clean or replace the tip
One hot streak with a misty, fine spray	Orifice partly blocked at the centre, breaking up the sheet	Inspect the pattern against a neighbouring nozzle	Clean or replace; check the filter that should have caught the particle
Intermittent clogging across the whole row	Filtration breakthrough: torn element, bypass open, tank debris stirred up	Inspect filter elements; check tank level and suction screen	Repair filtration; clean the tank; flush the header before restart
Clogging repeats in the same lane every few weeks	Corrosion products from one branch; scale in a dead leg	Sample the water at that header; inspect the branch	Flush or replace the branch; add scale/corrosion control
Spray angle shrinks but flow seems unchanged	Partial blockage of the edge-forming surfaces inside the tip	Compare the wetted width of suspect vs good nozzles	Clean or replace; inspect the internal slot profile

Cleaning discipline matters. A precision-ground tip should never be rodded out with a steel wire: the wire enlarges or scores the orifice, turning a clogging problem into a wear problem. The correct sequence is: remove, soak in the descaling or solvent bath for your water chemistry, rinse, blow through with clean air. If a tip re-clogs twice, fix the filtration upstream instead of cleaning a third time.

Nozzle Wear: The Silent Flow Drift

Wear never looks like a fault. Erosion widens the orifice and rounds the internal edges, so a worn nozzle delivers more flow at the same pressure, not less, while its pattern widens and drops coarsen. Nothing in the control room changes: the pressure regulator holds its setpoint while every worn nozzle drifts off spec.

Symptom	Likely cause	Quick check	Remedy
Strip runs cold with no other change	Worn nozzles delivering excess flow; the zone is overpowered	Bucket-test: flow above the stamped rating means wear	Replace; fit harder orifice materials in abrasive service
Pattern looks broad and weak	Orifice edge erosion widening the fan and softening intensity	Compare wetted width against a new tip of the same spec	Replace; check whether abrasive solids are bypassing filtration
Flow readings creep up over months	Progressive orifice erosion	Compare current flow tests with the previous inspection record	Plan replacement by throughput, not by calendar alone
One lane cools more than its neighbours	One tip older or harder-worn than the rest of the row	Compare that lane’s flow against the row average	Replace the outlier; standardize rotation intervals
Drops visibly coarse at the same pressure	Worn internal surfaces atomize less effectively	Visual comparison with a new tip	Replace; review water hardness and abrasive loading

Replace nozzles on a throughput-driven schedule, litres passed or operating hours, not on visible failure, because by the time a worn nozzle looks bad it has been off-spec for thousands of tonnes. A flow-rotation scheme is standard: bucket-test a sample at each stop, replace any nozzle beyond plant tolerance, and record the numbers so the next troubleshooting session starts from history instead of from zero.

The Physics Behind Every Fix

Each fix in the tables comes from four physical relations; internalizing them lets you predict faults before they appear:

Flow scales with the square root of pressure (Q ∝ √P). Doubling pressure raises flow about 41%, which is why “crank up the pump” fixes nothing when the real problem is a clogged filter.
Coverage width is geometry, not pressure. Wetted width is twice the standoff times the tangent of half the spray angle: at a 300 mm standoff a 25° fan covers roughly 130 mm, a 40° fan roughly 220 mm. Raising pressure only intensifies within the same width.
Heat removal is mass flow times temperature rise (Q = m·cp·ΔT), or latent heat when water evaporates. If the strip runs hot at nominal flow, the water is too warm, not staying in contact, or not reaching the strip.
Film boiling blocks contact. Above a surface temperature threshold, water flashes a vapour film that insulates the steel; a high-impact spray breaks it. That is why high-pressure flat fans exist for the hottest zones and a fine mist, excellent at moderate temperatures, is useless on a red-hot strip.

These relations are also your acceptance test: after any change, the temperature profile should move the way the physics predicts, and the gauge, flowmeter and bucket test should agree. When they contradict each other, the measurement is usually wrong.

The Field Inspection Procedure

Strip cooling troubleshooting works best as a fixed sequence, because each step gathers data and rules out causes. Run it in this order:

Confirm the problem is real and recent. Pull the logged temperature profile and compare with previous coils; note line speed, width, grade, and water temperature.
Check the water supply first. Supply temperature, tank level, pump discharge, filter differential pressure, before touching any nozzle. Upstream faults are the cheapest to fix.
Read every header pressure. Read the gauge at every cooling zone, not just the suspect one; a zone far below its neighbours is the fault zone.
Map the temperature across the width. Use a handheld pyrometer across the strip at each suspect zone exit, or the line’s own array. Mark cold and hot lanes on a header sketch. They line up with nozzle positions.
Inspect the suspect lanes. Spray shaped like its neighbours, tip not blocked, correct angle and gasket, no mixed tip types.
Bucket-test the suspects. With the zone isolated, measure flow at design pressure from the fault lane and a healthy lane. A nozzle outside plant tolerance is replaced regardless of how it looks.
Verify the fix. Restart and watch the temperature profile return to its pre-fault shape at the same speed and grade. If it does not move, you fixed the wrong thing. Go back to step 3.

Steps 1–4 take less than an hour and cost nothing, yet they identify the cause family in most cases. The common mistake is skipping to step 5, pulling nozzles off at random, which replaces a data-driven repair with a parts lottery.

The Maintenance Plan That Prevents Most Strip Cooling Faults

Prevention is cheaper than diagnosis, and most of these faults are preventable with a plan built around four pillars:

Water quality. The biggest driver of clogging and wear. Maintain filtration, change filter elements on a schedule or set differential pressure, keep the recirculation tank clean, and treat for your scale and corrosion chemistry.

Nozzle inspection and rotation. Inspect a nozzle sample at every stop, biased toward the highest-velocity zones, bucket-test against the stamped rating, and replace any outside tolerance. Rotate on a throughput-based interval and standardize the row to one type and angle.

Header and supply checks. At each stop, verify header level and alignment, gauge calibration, filter condition, and pump performance. Air-locked headers and leaking zone valves are found here, before they become temperature faults.

Records and baselines. Log temperature profiles, header pressures, nozzle flows, and every change. When a fault appears, “what changed since the last good coil” is answered in minutes from the records.

A workable rhythm: daily logs of supply temperature and header pressures; weekly filter-differential and row-walk checks; monthly bucket-tests and a temperature-log review; full service at each stop. Run on that rhythm, the fault families show up first in the records, a small pressure trend, a slightly wide spread, long before they reach a coil.

Worked Example: What the Numbers Actually Say

A worked calculation shows how fast the physics separates causes. Suppose a zone must remove 1 MW from the strip with 10 °C of usable water temperature rise.

Sensible cooling only: Q = m·cp·ΔT, so m = 1000 kW ÷ (4.18 kJ/kg·K × 10 K) ≈ 24 kg/s: about 1,430 L/min that must actually contact the strip and rise 10 °C.
Evaporative cooling: each kilogram that fully evaporates absorbs about 2.26 MJ, so the same 1 MW needs roughly 0.44 kg/s ≈ 26 L/min evaporated: about fifty times less water, provided drops are small enough to evaporate or the surface hot enough to flash them.

That is why the first question in strip cooling troubleshooting is never “which nozzle” but “is the pressure, the water temperature, and the contact really what the gauge says”.

Frequently Asked Questions

What is the first thing to check when strip cooling becomes uneven? The supply side: water temperature, tank level, filter differential pressure, and every header pressure gauge, not just the suspect zone. Most uneven-cooling faults trace back upstream of the nozzles.

Why does edge overcooling cause wavy edges? Edges lose heat through the side face as well as the top and bottom, and run-off removes extra heat. The edge ends up harder and less ductile than the middle, so it responds differently to the same downstream strain and develops edge wave.

How do I tell a clogged nozzle from a worn nozzle? By flow and pattern direction. A clogged nozzle sprays less, often with a distorted or misty pattern. A worn nozzle sprays more than its stamped rating, with a wider, weaker pattern. A bucket test at design pressure separates them unambiguously.

How often should strip cooling nozzles be replaced? On a throughput-based interval, operating hours or litres passed, not on a calendar or visible failure. Bucket-test a sample at each stop, replace anything outside tolerance, then set the rotation interval from the observed drift rate.

Why is a fine mist nozzle wrong for the hottest zones? Above the film-boiling threshold the surface is insulated by a vapour layer, and a low-impact mist cannot break it. The water never touches the steel. High-impact flat fans punch through that film.

Printable Troubleshooting Checklist

Summary: Troubleshoot the Water, and the Steel Follows

Strip cooling troubleshooting is a search for where the water is not doing what the design says it should. The five fault families each have a signature pattern and a short list of causes, and the physics (square-root flow, geometric coverage, sensible and latent heat, film boiling) explains every fix. Read the pattern, check the supply side first, bucket-test the suspects, verify the fix, and keep the records, because the coil is only as good as the water distribution that made it.

If your strip cooling section needs new flat fan nozzles for a replacement rotation, or you are designing a new cooling zone and want to check geometry and overlap before you buy, contact us with your duty, header pressure, standoff, strip width and water supply, and we will help you size the row. For which nozzle family belongs in which cooling zone, see our guide to cooling nozzle selection, and for how wear quietly inflates flow, read about nozzle wear and silent over-application.