Back to FWG & Hydrophore
Auxiliary Machinery & Shipboard Systems

Fresh Water Generator — Vacuum Boiling, Operation & Faults

How 92% vacuum lets engine waste heat boil seawater at 50 °C, and what high salinity or low output is telling you.

12 min read
Intermediate
Auxiliary Machinery & Shipboard Systems
Key Principles at a Glance 6 points
  • Vacuum is the whole trick: near-92% vacuum drops the boiling point so ~80 °C jacket water evaporates seawater at ~50 °C shell temperature.
  • Vacuum is also the most fragile thing — any air leak or sick ejector kills output before anything else breaks.
  • Roughly half the feed becomes distillate; the ejector must continuously remove the other half as brine or density climbs and carry-over follows.
  • High salinity points at feed rate, evaporation temperature, brine level, leaking tubes or a dirty probe; low output points at scale, vacuum, ejector, jacket heat or the distillate pump.
  • Low-temperature distillate is never sterile and never made near coasts — chlorination, electro-katadyn or UV plus offshore-only production.
  • Reverse osmosis is the rival method: seawater forced through spiral-wound membranes, output set by pressure, temperature and salt load.

1. Why Vacuum Turns Waste Heat Into Fresh Water

Idea in one line: lower the pressure and water boils cooler — so engine jacket water too cool to boil anything on deck becomes a boiler inside a vacuum shell.

At atmospheric pressure seawater boils near 100 °C and jacket water at ~80 °C could never touch it. Pull the shell down to ~92% vacuum and the boiling point collapses to ~50 °C — suddenly that same jacket water is a strong heating medium. Seawater evaporates, salt stays behind in the brine, and the vapour condenses on seawater-cooled tubes as distillate. Heating medium is jacket water or steam; the unit is a self-contained vacuum package.

MORE VACUUM → LOWER BOILING POINT 100°C 50°C vacuum → atm: boils ~100°C 92% vacuum: boils ~50°C jacket water ~80°C now boils it
~92%Vacuum to build before feeding
~50 °CShell boiling temperature
≥ 3 barEjector seawater pressure
~½ feedBecomes distillate; rest is brine

What lives inside the package: coated mild-steel shell, vertical-tube evaporator bundle, shell-and-tube condenser, brine-and-air ejector with its pump, distillate (condensate) pump, vacuum gauge, distillate level gauge, salinometer, dosing connection and a shell relief valve. One saturation pressure rules both chambers — evaporation and condensation sit at the same boiling point for that vacuum.

2. Start-Up — Vacuum First, Heat Last

Idea in one line: the start sequence builds the vacuum trap before letting any heat or feed in — feed without vacuum just fills the shell with salty water.

ejector≥ 3 bar vacuum~92% feed +jacket heat salinometer on,distillate to tanks
1

Ejector lines open, pump started, ≥ 3 bar proved — suction, discharge and overboard valves open first. Why: the air ejector is the vacuum pump; weak motive pressure means no vacuum ever builds.

2

Wait for ~92% shell vacuum on the gauge — feed nothing until it arrives. Why: no vacuum, no low-temperature boiling, no water.

3

Crack the feed valve, set feed pressure and flow — restricted feed, not full bore. Why: only about half the feed can flash; flooding the chamber drowns boiling and pushes brine toward the demisters.

4

Admit jacket water gradually, vent the evaporator top until solid — open inlet and outlet slowly, crack the air vent until water-filled. Why: sudden heat shocks tubes and trapped air blankets heat transfer.

5

Salinometer panel on, distillate pump primed, product to nominated tanks — watch the suction sight-glass for solid prime. Why: an air-locked distillate pump cavitates while off-spec water could already be heading for the tanks.

3. Running Watch — What Each Reading Means

Idea in one line: every gauge is an early warning for one failure — flashing stops means heat or vacuum went, brine rising means the ejector stopped clearing.

SHELL SIDE flashing in glass ✓~50°C · >90% vacbrine level mid WATER SIDES condenser in/out ΔTejector > 3 bardistillate press + flow QUALITY salinity readingdosing level/flowpump amperes ×2
CheckNormalDrift means
Flashing in shell glassLively boiling visibleDead glass: vacuum lost or jacket heat gone
Brine levelSteady, neither high nor lowHigh: ejector nozzle clogged or overboard shut; low: feed starved
Shell temperature / vacuum~50 °C, vacuum above 90%Temperature up with vacuum down: air leak — the single most important reading
Condenser seawater in/outSensible temperature riseNo rise: over-cooled evaporator or flooded condenser
Ejector, distillate, dosing, ampsEjector > 3 bar; steady distillate flow; dosing flowing; normal currentsSagging ejector pressure, swinging amps or empty doser precede the fault tables below

4. Plate vs Flash — Where Every Stream Goes

Idea in one line: both designs flash seawater into steam and scrub the salt spray out before condensing — plate type boils in place, flash type boils by falling pressure.

PLATE: BOIL IN PLACE FLASH: BOIL BY PRESSURE DROP feed (restricted) → boils½ steam ↑ · ½ brine ↓ separator + demister → condenserdistillate via salinometer heat chamber 1 →lower-pressure chamber 2 flashes to steam on entrycondenses on cool tubes

Plate type in detail: hot jacket water crosses the heated nest while restricted seawater feed enters the low-pressure chamber and boils — about half flashes to steam, half stays as brine. Steam crosses a separator that knocks out water particles, sheds released air and non-condensable gases to the air ejector, passes baffles and demisters that stop saline carry-over, condenses on the cold bundle, and the distillate pump delivers it past the salinometer. Brine collecting at the bottom is drawn off continuously by the brine ejector — stop removing it and density climbs until priming carries salt into the product.

Scale chemistry: heating drives gas out of calcium bicarbonate, leaving calcium carbonate that plates tubes as soft scale. At ~60 °C coil duty it stays manageable — push temperatures up and it hardens into an insulating blanket that strangles output.

Flash type in detail: seawater heats in one chamber, then spills into a second chamber at substantially lower (sub-atmospheric) pressure — the sudden fall flashes part of it to steam without further heating. That steam condenses on much-cooler salt-water tubes exactly as above, and baffles plus demisters guard the distillate in both designs. Single or two-stage layouts exist.

Plate FWG flow diagram with jacket circuit, evaporator, condenser, ejectors and salinometer loop
Figure 1: Trace every stream — jacket heat in, feed flashing, brine out through the ejector, vapour through separator and demisters, distillate past the salinometer. Vacuum holds the whole play together.

5. Two Faults, Two Checklists

Idea in one line: salty product means salt is getting through or the sensor is lying; low output means heat, vacuum or pumping is choked — work each list in order.

SALTY? → SALT PATH OR SENSOR LITTLE WATER? → HEAT/VACUUM/PUMP feed ↓ · cooling ↓brine level · tubes · probe scale · air leak · ejectorjacket temp · distillate pump cheapest first: cut feed,ease cooling, clear brine,then plug tubes, clean probe vacuum first always:seal leaks, inspect ejector,clean scale, raise jacket heat
High salinity — check in orderWhy it salts the product
Capacity too high — reduce feed rateOver-firing floods the separator; spray primes straight into the steam lane
Evaporation temperature too low — reduce seawater coolingOver-cooled shell condenses poorly and drags brine droplets along
Brine level too high — ejector nozzle clogged or overboard valve shutRising brine drowns the vapour space; open the overboard path and clear the nozzle
Condenser tube leaking — find and plugSeawater leaks directly into distillate inside the condenser
Salinometer probe defective or dirty — clean or renewA fouled probe condemns good water or passes bad; prove the instrument before opening the plant
Low capacity — check in orderWhy output falls
Heater tubes fouled with scale — cleanScale insulates the bundle; jacket heat never reaches the seawater
Air leaks, low vacuum — find and sealAir raises the boiling point and blankets condensation; vacuum first, always
Defective ejector, low vacuum — inspect and renewWorn nozzle or weak motive pressure cannot hold vacuum or clear brine
Jacket water too cool — raise temperatureSlow-steaming or cold engine starves the evaporator of heat
Defective distillate pump — check pressure and motor loadCavitation or wear leaves made water stranded in the condenser

6. The Rival Method & Drinking-Water Rules

Idea in one line: osmosis can be forced backwards with pressure — but membranes, not vacuum, become the fragile heart of the plant.

salt waterforced →high pressure membrane pure waterpassessalts rejected spiral hollow fibresoutput ∝ pressure, tempoutput ∝ 1 ÷ salt load

How it works

Natural osmosis pulls pure water toward the salty side. Forcing salt water against a semi-permeable barrier reverses it — light water molecules pass, large dissolved salt molecules cannot. The barrier is hollow fine fibres spirally wound into cartridges, one-way by construction.

What sets output

Net drive pressure across the membrane, pressure loss through it, feed-water temperature, and how salty the feed is. Membranes must survive high pressure, and big production needs big membrane area.

Drinking-water law

Low-temperature evaporation never sterilises. Never generate near coasts where micro-organisms load the feed — sterilise every drinking batch by chlorination, electro-katadyn or ultraviolet, dosed and monitored to current notices.

Never near shore, never unsterilised: coastal water carries the organisms vacuum boiling cannot kill. Offshore production plus sterilisation is not optional — it is the licence to drink.