Rotary Jet Head Cleaning of Marine Cargo and Slop Tanks

A cargo tank that returns from the wash berth with heel in its frames costs a ship a loaded slot. Off hire on a coated chemical tanker runs 12,000 to 28,000 USD per day (published range). A failed gas free test adds 8 to 20 hours of waiting. A class survey finding on a slop tank can hold a vessel at anchorage for another day. Most of these failures trace back to coverage, not to nozzle wear.

This case study follows a composite owner rebuilding its marine tank cleaning equipment around fixed rotary jet heads. The story below is a generic composite. No plant record is reproduced and no real company is named. The arithmetic is shown in full, so a technical buyer can recompute every step.

Case Snapshot

Parameter	Manual portable lance (baseline)	Fixed rotary jet heads (retrofit)
Vessel and tank class	Coated chemical tanker, 12,000 to 20,000 m3, 8 cargo tanks	Same class, 8 cargo tanks plus 2 slop tanks
Tank size	2,000 m3 per cargo tank, r of 6.0 m, h of 17.7 m	Unchanged
Soil type	Vegetable oil, fatty acid, glycol, light fuel oil, slop emulsion	Same product set
Head count	2 hand held lances, one crew pair	1 bottom mounted and 2 wall mounted heads per cargo tank, 2 per slop tank
Flow per head	60 to 80 L/min at 8 bar	694 L/min at 7 bar, rated 700 L/min
Pressure	8 bar, or 116 psi	7 bar, or 101.5 psi
Cycle time per tank	330 minutes	55 minutes including drain
Water per cycle per tank	15.0 m3	6.3 m3
Evidence retained	Crew log and visual check	Coverage test records, gas readings at 3 points, class survey sign off

The figures in this snapshot describe the generic shape of this class of job, not a plant record.

The Challenge

A coated cargo tank is not a smooth vessel, and the road-going cousin of this duty, the ISO tank container, fails the same audit between cargoes; the depot-side sizing sits in the ISO tank cleaning guide. Internal framing, brackets, ladder rungs and heating coils add 40 to 100 percent to the wall area. A 2,000 m3 tank looks simple on a general arrangement drawing. Its internal surface is really about 900 to 1,400 m2 (published range for framed tanks). Every square metre must see an impinging jet at useful pressure.

Jet impact decays quickly with distance. A 7 bar orifice gives about 2.2 bar of impact at 2 m of throw. At 6 m it gives 0.6 bar, and at 12 m about 0.2 bar (published range, square law decay).

Soil removal needs 0.5 to 1.0 bar at the target surface (published range). The usable radius is therefore nearer 6 m than the 12 m reach on a datasheet. Coverage, not reach, decides whether a tank passes.

Manual lancing has two failure modes. The crew cannot reach behind the transverses without entering the tank again and again. Every entry is a permitted enclosed space entry under the vessel safety management system. The second failure is repeatability. Two operators on two watches leave two different coverage patterns on the same steel.

The cost lands on the voyage. A portable lance round costs 11 man hours per tank. At 45 USD per man hour (published range) that is 495 USD per tank in labour alone. Water, heating, slop disposal and the gas free wait sit on top of that figure. A 12,000 m3 chemical tanker with 8 cargo tanks spends about 88 man hours on a full round.

Gas free is the harder gate. Residue in frames releases hydrocarbon vapour slowly after the wash. A tank that reads clean at the hatch can drift back above the threshold overnight.

SOLAS hot work and entry requirements demand readings below the agreed vapour limit before a permit is issued. NFPA 30 governs the same questions alongside a regulated berth. IMO MARPOL Annex I controls what may be discharged and where. The IMO resolution on tank cleaning equipment sets the machine duty a class society expects.

Class survey rejection ends that chain. A surveyor who cannot see coverage records will not close the item. Three open findings across five attended tanks is a common pattern on an audit (published range). The remedy is not more water. It is a layout that produces evidence as a by-product of the wash.

Coverage Geometry of a Three Arm Rotary Head

The heads here are generic 3 arm fully rotary turbine driven jet heads. Published ranges for the family are 200 to 1,600 L/min flow, 5 to 10 bar pressure and 3 to 12 m impact radius. Nothing depends on a maker part number. The geometry is what matters.

Each arm carries four orifices, angled forward and back along the axis. The reaction force of the jets turns the arms. A turbine wheel inside the housing adds torque, so the arms keep turning at low flow. Rotation speed sits at 0.5 to 3 rpm (published range).

Full rotation means the head sweeps every vertical line in reach once per revolution. The horizontal angle of each nozzle sets the vertical band that nozzle covers. The coverage map is built on a grid. Divide the wall into panels, then mark each jet passage across one revolution.

Overlap is the working margin. A map with 20 to 30 percent overlap tolerates small errors in head placement. A map with no overlap fails when one arm runs slow or one orifice erodes.

The exchange index follows from the same geometry. Divide tank volume in cubic metres by head flow in litres per minute. The deck shorthand reads 2,000 divided by 700, which is 2.9 minutes per volume exchange. Real cycles need 3 to 8 volume exchanges (published range). That gives 8.6 to 23.2 minutes of jet run time per tank.

Converted strictly, 2,000 m3 is 2,000,000 L. One full liquid exchange at 700 L/min would take 2,857 minutes. That is why these machines recirculate the tank inventory instead of displacing it.

Profiled Surfaces and Shadow Zones

Framing creates shadow zones. A transverse web, a bracket toe, a pipe run and a bellmouth hide the steel behind them from a straight jet. The layout duty is to shorten those shadows, not to claim they vanish.

Three design moves reduce shadowing. Nozzles are tilted 10 to 20 degrees off the arm axis, so the jet wraps past a frame edge. Heads are placed so at least two see any given panel. The final rinse uses a portable lance on the 5 to 10 percent of area no fixed head reaches.

Wall Mounted and Bottom Mounted Heads Compared

A bottom mounted head sits in a sump or deck penetration on the tank centreline. Its vertical axis suits floor work, and it does not obstruct cargo handling. Its weakness is throw. The jet must cross the full tank height to reach the deckhead. Impact at the far wall is then the weakest point on the map.

A wall mounted head sits 2.0 to 3.0 m above the tank bottom. It fixes to a welded pad or a certified bracket. Throw distances shorten, so impact on side walls and deckhead improves. The cost is a new tank penetration with an approved weld detail and a guard against cargo gear.

The combination wins in this class of tank. One bottom mounted head covers the floor and lower strakes. Two wall mounted heads cover the mid body, deckhead and framing. Three heads per 2,000 m3 tank keep the longest throw under about 7 m. That sits inside the 0.5 to 1.0 bar window.

Placement must respect class society rules and the IMO resolution on tank cleaning equipment. The machine must be rated for the tank, and the penetration detail must be approved before welding. On a gas free tank the head sits in a controlled atmosphere. Its electrical equipment must carry the correct ATEX zone marking for the condition of use.

Cleaning of Profiled Surfaces With Rotating Nozzles

Cleaning of profiled surfaces with rotating nozzles works through repetition, not one long soak. Each point of steel receives several short impingements instead of a single soak. A pass at 0.5 to 3 rpm delivers a short pulse to each panel. Time, pressure and repetition do the work.

The number of impingements per point is a design output. At 2 rpm a head turns once every 30 seconds. A 20 minute cycle gives 40 revolutions, so every line in reach is swept 40 times. That is far more than the 3 to 8 volume exchanges the liquid side needs.

Orifice wear changes the map slowly. A 5.8 mm orifice that erodes to 6.1 mm passes more flow at the same pressure. It also shifts the spray pattern outward. A filtration and inspection routine keeps erosion inside the margin the 20 to 30 percent overlap allows. The selection logic behind these drills is set out in our guide on tank cleaning nozzle selection.

The Solution

The marine tank cleaning equipment layout is fixed and simple. Each cargo tank carries one bottom mounted head on the centreline and two wall mounted heads at 2.4 m. Each slop tank carries two wall mounted heads. Eight cargo tanks and two slop tanks need 24 heads, run in banks of four.

The pump set is sized on the bank, not the tank. Four heads at 694 L/min each give 2,776 L/min at 7 bar. That is 166.6 m3/h. A variable speed drive trims the bank to the duty of the tank in service. A slop tank then never runs at full cargo flow.

Filtration is the reliability item. A duplex strainer at 100 to 200 microns sits upstream of each bank. Differential pressure gauges and a changeover valve let the crew clean one basket while the other works.

Slop routing uses gravity and separation. Wash water drains to the tank sump and transfers to a dedicated slop tank. The slop tank decants free oil at the top and clarified water below. Recovered water is filtered, reheated and returned to the wash circuit. The fixed heads come from the same heavy duty rotary jet head family used on land based tank cleaning.

Controls are deliberately plain. A PLC runs a fixed cycle of pre-rinse, hot wash, rinse and drain. Cycle counters and pressure logs export to the vessel planned maintenance system, so the evidence pack is ready before the survey.

The Arithmetic Behind a Wash Cycle

Orifice flow follows Q equals 0.65 times d squared times the square root of pressure. Put a 5.8 mm drill on 7 bar. Root seven is 2.646, so the sum is 0.65 times 33.64 times 2.646. That gives 57.9 L/min per orifice. Twelve orifices give 694 L/min, called a 700 L/min duty head.

Orifice count scales flow at fixed pressure, because total area scales. Twelve orifices of 5.8 mm hold 12 times 26.4 mm2. That is 317 mm2 of open area. Doubling to 24 orifices doubles area and flow to about 1,388 L/min. That is why one housing family spans 200 to 1,600 L/min (published range).

Tank volume follows V equals pi times r squared times h. For r of 6.0 m and h of 17.7 m, V is pi times 36 times 17.7. That is 2,002 m3, rounded to 2,000 m3 or 2,000,000 L. Washing time is V divided by Q at the stated coverage assumption. That gives 2,000 divided by 700, or 2.9 minutes per volume exchange on the deck convention.

Real cycles run 3 to 8 volume exchanges (published range). So 3 times 2.9 is 8.6 minutes, and 8 times 2.9 is 23.2 minutes of jet run time per tank.

Pump power follows kW equals Q in L/min times P in bar divided by 612. One head at 700 L/min and 7 bar draws 700 times 7 divided by 612. That is 4,900 divided by 612, or 8.0 kW of hydraulic power.

At 75 percent overall efficiency the shaft power is 8.0 divided by 0.75, which is 10.7 kW. At 60 percent it is 8.0 divided by 0.60, which is 13.3 kW. The published range is 11 to 13 kW of shaft power per head.

A bank of four heads needs 4 times 8.0, which is 32.0 kW hydraulic. That is 43 to 53 kW at the motor shaft, or about 58 to 71 hp at 0.746 kW per hp. Motor sizing follows the shaft figure, not the hydraulic one.

Heat load for a hot wash follows a simple rule. kW equals mass flow in kg/s times cp of 4.18 kJ per kg K times delta T in kelvin. Water at 700 L/min is about 700 kg/min. Divide by 60 and the mass flow is 11.67 kg/s.

With a delta T of 40 K the load is 11.67 times 4.18 times 40, which is 1,951 kW. That is 2,615 hp. This is the dominant electrical and boiler load in the plant, so hot wash runs on a reclaim batch, not the full stream.

Water reclaim follows before minus after, divided by before, times 100. A tank that used 15.0 m3 before and 6.3 m3 after shows 8.7 divided by 15.0. That is 0.58, or 58 percent. Across 8 cargo tanks the round drops from 120.0 m3 to 50.4 m3.

In energy terms, heating 6.3 m3 through a 40 K rise takes 6,300 kg times 4.18 times 40, which is 292.6 kWh. The same warm wash on 15.0 m3 needs 696.7 kWh.

Unit conversions run through every step. One m3 is 1,000 L, 1 L/min is 0.06 m3/h, 1 bar is 14.5 psi and 1 hp is 0.746 kW. A 7 bar wash is 101.5 psi. A 694 L/min head is 41.6 m3/h. A 10 bar ceiling is 145 psi.

Orifice sizing at these duties is covered in the guide to high pressure water nozzle orifice selection.

Slop Handling and Water Reclaim

On a chemical tanker the slop tank is not just a drain. Washings can carry residue that IMO MARPOL Annex I restricts for oily mixtures. The IBC Code tracks the same cargo through the record book. Decanting separates free oil at the top from clarified water below. The volume sent for shore disposal then falls.

Reclaim quality is a specification, not a hope. The circuit uses 100 to 200 micron filtration plus a coalescer. The final rinse runs on fresh water from the technical line.

The water budget follows the reclaim ratio. At 15.0 m3 per tank the fleet needed 120.0 m3 for a round of 8 cargo tanks. At 6.3 m3 per tank the same round needs 50.4 m3. Slop generation tracks the same direction, from 15.0 m3 to 6.3 m3 per tank. That cuts both the heating load and the shore disposal fee.

Slop tank size stays at the class approved figure. The gain is in the number of transfers, not tank volume. Fewer transfers mean fewer pump runs and less vapour release. The same reclaim logic appears on land in road tanker and IBC wash bays, where the only difference is the size of the slop inventory.

Gas Free, Hot Work, and Entry Acceptance

Gas free acceptance is a set of readings, not an opinion. A portable gas tester is run at 3 points in the tank: top, middle and bottom. Vapour stratifies, so a single hatch reading can miss a pocket.

Hydrocarbon vapour must sit below 10 percent of the lower explosive limit for entry. Hot work needs below 1 percent of the limit (published threshold). Oxygen must read 20.5 to 21 percent. Toxic gas must stay below the workplace exposure limit.

Ventilation runs at 8 to 12 air changes per hour (published range). SOLAS hot work and gas free requirements drive the permit. Each reading is logged with time, position and instrument serial number.

The retrofit shortens the wait. Manual washing left standing water in the sump and in frame roots. Venting then took 6 to 10 hours before readings held.

Fixed heads drain the sump harder and run a controlled final rinse. Clearance falls to 2 to 3 hours. On a ship waiting for a berth, that gap decides whether it catches a tide.

ATEX zone language matters at the permit stage. A tank under wash with residual vapour can be treated as a hazardous area. Any portable equipment used at the hatch carries the right zone rating.

Class Survey Evidence

A class surveyor wants a record, not a reassurance. The retrofit produces three. The coverage test record shows the map, the overlap and the measured throw. The pressure and flow log shows each bank held its duty across the cycle. The gas free log shows the 3 point readings against the threshold.

The coverage test is run wet. The crew mark a grid on the wall and run one full cycle. They then inspect for dry or lightly wetted panels. Overlap is judged against the 20 to 30 percent target.

Findings close faster when the evidence is in one pack. On the baseline round the owner closed 3 of 5 attended tank findings. The rest were deferred for a further wash. After the retrofit the same fleet closed 5 of 5. Each tank arrived at survey with a complete coverage and gas free pack.

Class society rules also govern the penetration and the welded pad. The pad detail, weld procedure and pressure test all sit in the approved documentation.

The Results

Metric per cargo tank	Manual lance (before)	Rotary jet heads (after)	Direction
Cycle minutes	330	55	down 83 percent
Water per cycle	15.0 m3	6.3 m3	down 58 percent
Hot water or steam load	696.7 kWh	292.6 kWh	down 58 percent
Man hours per tank	11.0	1.5	down 86 percent
Gas free clearance time	6 to 10 hours	2 to 3 hours	down about 65 percent
Survey findings closed	3 of 5	5 of 5	up
Per tank cost direction	baseline	down 30 to 45 percent (published range)	down

The table holds the same tank, the same product set and the same class rules. Only the wet end of the marine tank cleaning equipment changed. Holding everything else constant is what makes the comparison readable.

Reading the Numbers Against Published Ranges

Each figure in the results ties to a published band for this equipment class. Flow sits at 694 L/min inside the 200 to 1,600 L/min family range. Pressure sits at 7 bar inside the 5 to 10 bar range. Effective impact radius sits at 6 to 7 m inside the 3 to 12 m geometric range. The 0.5 to 1.0 bar soil removal threshold closes the outer band.

Power and heat follow the same check. Shaft power of 10.7 to 13.3 kW per head matches the 11 to 13 kW band at 60 to 75 percent efficiency. The 1,951 kW hot wash load sits near the top of what one head can demand. That is why a bank is staged rather than run hot in full. Water reclaim of 58 percent sits inside the 40 to 70 percent band for closed loop systems (published range).

Time figures need one caution. The 2.9 minute exchange index is a deck shorthand, not a wet clock. The 55 minute cycle includes pre-rinse, hot wash, rinse, drain and physical handling. No formula captures the handling. Anyone who quotes 8.6 to 23.2 minutes as a full cycle has skipped the parts that consume the day.

Cost direction of 30 to 45 percent is net of reclaim and staging gains but before capital recovery. A retrofit pays back on labour and gas free hours alone across a normal drydock span. The number to watch is the closure rate on survey findings. A deferred finding repeats the whole cost.

Why This Matters

Every fleet owner runs the same argument. Manual lancing looks cheap because the labour is already aboard. The true cost hides in off hire, repeat cycles and survey deferrals. Those items appear far from the wash berth on the ledger.

Standardising marine tank cleaning equipment on the wet end changes three things. Spares and nozzle drills become one line item across the fleet. Crew competency transfers between ships, because the cycle and the coverage map are the same. The evidence pack becomes routine instead of an emergency.

The same discipline applies beyond cargo tanks. The reclaim and filtration logic carries to other wash duties. The nozzle level decisions echo land based work such as the railcar wash nozzle boom retrofit and the wider tank vessel cleaning method set. The marine case is simply the version with the tightest penalty for getting it wrong.

Next Steps For Your Fleet

If your tanks fail gas free or close the survey late, the layout and the arithmetic are fixable. Start with a coverage map, a bank flow figure and your gas free acceptance criteria. Bring those three items and we will size the heads, the pump set and the reclaim loop. Send the tank drawing, the product list and the current cycle time.

We will return a head layout, a pump and filtration schedule and a water and heat budget you can take to class. The same marine tank cleaning equipment logic applies to a single slop tank or a fleet wide standard. If you want to compare hardware before we scope the retrofit, review the tank cleaning nozzle range. Get in touch through the contact page to start the sizing.