Ship's Log
Places, passages and repairs from aboard SV Eelyos.
- Engine Oil Check Maintenance
Checked the oil this morning after yesterday's run down to Tenakee — good colour and right level.
This was the first proper run (36 nm, 5h36m under power, Swanson Harbour → Tenakee Springs) since the severe diesel dilution that forced the emergency oil change on 2026-08-05, and the fuel lift pump, injector, valve lash and head bolt work that followed. No sign of the dilution coming back.
Inspection only — nothing added, nothing changed.
- Torque Cylinder Head Bolts Maintenance
Tightened cylinder head bolts to spec. Done alongside the injector replacement and valve lash adjustment.
Torque
105 ft·lb, per the documented service manual figure.
Torqued HOT — engine run under load for 40 minutes immediately beforehand, so the head was at true working temperature rather than merely idled warm. Retorquing these engines hot is the standard procedure.
Corroborated: published Ford Lehman 4-cyl 80 hp figures quote 105–110 ft·lb, so 105 sits at the bottom of the accepted range.
Sequence was correct
Head bolts torqued before the valve lash adjustment. Retorquing pulls the head down and changes rocker geometry, so lash must be set afterwards or it will be out. Both jobs shared the same 40-minute under-load warm-up, so the head and valvetrain were at the same working temperature for each.
Not recorded
- whether the specified tightening sequence (centre-outwards, staged steps) was followed
Interval
Confirmed every 400 engine hours (2026-08-06), as one of a four-item 400 h service group: valve lash, injector replacement, cylinder head bolt torque, and a raw water pump check. Schedule now set — next due at 7,594 h.
Supersedes the 500–1000 h figure previously noted here from forum / owner's-manual sources.
- Replace Fuel Injectors Maintenance
Replaced the fuel injectors with new units from Harbor Marine, Everett WA. Full set.
This is the repair for the crankcase diesel dilution investigated 2026-07-11 → 2026-08-05.
Confirmed source
Fuel escaping at the top nut of the rearmost (No.4) injector, dumping the entire leak-off circuit into the rocker box and down to the sump at ~1.0 L/engine-hour — proven by zero return flow at the tank end with a verified-clear return line.
Replacing the whole set supersedes the copper-washer reseal that was the minimum fix, and removes the other three as future suspects. Reasonable at 7,194 h with a known-leaking set.
Still unconfirmed as fixed
Two checks, fastest first:
- Immediate — run ~2 min with the leak-off return routed into a jug. Fuel returning to tank = circuit restored. This is the direct inverse of the 2026-08-05 test that returned nothing.
- ~1 h run — confirm the oil level holds. At ~1.0 L/engine-hour any residual leak is unmistakable within a single hour.
New injectors were fitted with the engine oil freshly changed (2026-08-05 @ 7,192.84 h), so the sump is a clean baseline for check 2.
- Valve Lash Adjustment Maintenance
Valve lash (valve clearance) check and adjustment. Done after the cylinder head retorque, which is the correct order — retorquing pulls the head down and changes rocker geometry, so lash set beforehand would be wrong.
Clearances set
Per the documented service manual figures:
Valve Clearance Inlet 0.015 in Exhaust 0.012 in Set HOT — engine run under load for 40 minutes immediately beforehand, so it was fully warmed and heat-soaked at normal operating temperature, not merely idled to temperature.
Running under load matters: it brings the head, valve stems and pushrods to true working temperature, which idling does not, and lash changes with the thermal expansion of those parts.
These are the manual figures and supersede the unverified forum values previously noted here (quoted as cold settings of 0.014–0.018 in inlet / 0.010–0.014 in exhaust). No further re-check at temperature is needed — the adjustment was itself made hot.
Reference point for next time: hot, under-load warm-up, 0.015 in inlet / 0.012 in exhaust.
As-found: all valves loose
Every valve required adjustment — all had to be tightened in, i.e. clearances had opened up beyond spec across the board.
Uniform drift in one direction across every valve is the signature of accumulated valvetrain wear over a long interval — wear at the rocker pads, valve tips, pushrod ends and tappet faces all opens the gap — rather than any localised fault. A single valve wildly out would point at something specific (bent pushrod, worn rocker, or a valve receding into its seat, which moves clearance the other way); all of them moving together points at elapsed time.
There is no prior valve lash entry anywhere in this maintenance record, consistent with it having been a very long time.
Consequence of running loose
Excessive clearance reduces effective valve lift and duration, so the engine breathes worse and runs noisier (tappet clatter), and the valvetrain hammers itself with every cycle — accelerating the very wear that caused it.
Expect the engine to be quieter and to breathe better now. Not related to the diesel dilution (that was the No.4 injector leak-off seal), but it will have been costing efficiency.
Interval
Confirmed every 400 engine hours (2026-08-06), as one of a four-item 400 h service group: valve lash, injector replacement, cylinder head bolt torque, and a raw water pump check. Schedule now set — next due at 7,594 h.
- Change Fuel Pump Oil Maintenance
Changed the Simms Minimec injection pump cam-box oil. Was 91.6 engine hours overdue on its 50 h interval (last done 2026-07-07 @ 7,051.25 h; 141.6 h elapsed).
Diagnostic result — NEGATIVE
The old pump oil was healthy: no diesel smell, correct level, normal appearance.
That is a strong negative for the injection pump as the source of the crankcase diesel dilution. Fuel worn past the Minimec plungers collects in this cam case, so contamination there is the expected signature of that failure mode. Its absence largely eliminates both:
- the worn-plunger path
- the pump drive-seal path — fuel reaching the timing case via the pump would have to transit this cam box first
Level being correct rather than raised is corroborating: no accumulation.
Minor caveat
If the cam box drained continuously to the engine sump, fuel could in principle pass through without accumulating — but it would still leave a diesel smell, and there was none.
Consequence
With the lift pump inspected and undamaged (2026-08-05) and the injection pump now cleared, the leading candidate for the ~1.0 L/engine-hour running-proportional leak is the injector leak-off / return line inside the rocker cover — brazed joints that crack and drain return fuel into the valve gear and down to the sump.
Known Ford Lehman failure mode, invisible externally, carries plenty of volume, and not yet inspected.
Next action: pull the rocker cover.
- Change Engine Oil Maintenance
Emergency engine oil change — severe diesel dilution. Drained roughly 1.5 US gal (~5.7 L) more than was added at the 2026-08-03 change. Done the same day as the fuel lift pump replacement.
RESOLVED. The source was the No.4 injector leak-off / cap seal, confirmed 2026-08-05. Injectors replaced 2026-08-06. Full trail below.
The rate — the key measurement
Corrected per Theo, 2026-08-05.
The ~5.7 L entered the crankcase during the 5.6 engine hours of running, not gradually over the intervening calendar time. Sequence: fresh oil added at the 2026-08-03 change → engine started → ran 5.60 h → dipstick checked, grossly over-full and diluted.
So the leak is running-proportional at ~1.0 L per engine hour (~0.27 US gal/engine-h) — not a static siphon at rest.
Basis Implied rate Fits? Per engine hour ~1.0 L/h yes Per calendar hour ~0.10 L/h no Progressive failure
1.0 L/engine-h cannot have applied over the preceding 2026-07-11 → 2026-08-03 interval (131.5 engine hours; that rate would imply ~130 L). The earlier interval showed dilution but far milder, so the fault degraded progressively and then let go.
Sump capacity not recorded, so the dilution fraction is unknown — but 5.7 L is a large share of any sump this size, and the engine ran 5.6 h in that state.
Filter status
Not stated for this change. The filter in service dated from the 2026-07-11 change (137+ h) and had been through the entire dilution period. Update this note once confirmed.
Investigation
Lift pump — downgraded
Old pump inspected after removal: no visible damage, not a ruptured diaphragm. Hypothesis weak. See the Replace Fuel Lift Pump log for detail.
Engine identified
Ford Lehman 80 hp (2712E / 2704E series, 4-cyl), using a Simms Minimec inline injection pump. That changes the differential: inline pumps have a cam case that can drain fuel to the engine sump, a path rotary pumps do not have.
Differential — cheapest / most likely first
- Injector leak-off / return line inside the rocker cover. The return gallery runs under the rocker cover with brazed joints; a cracked joint dumps return fuel into the valve gear and straight down to the sump. Running-proportional, high volume, invisible externally. Known Lehman failure mode. Check first.
- Simms Minimec pump plungers worn, fuel passing into the pump cam case and on to the sump. Direct test: has the injection pump oil level risen, and does it smell of diesel?
- Injection pump drive seal (pump-to-gear) failing, fuel into the gear case then the sump.
- Lift pump — weak; old unit inspected and undamaged.
Field checks
- Hold the dipstick upside down and watch how fast the oil runs — thinned oil moves visibly faster.
- An oil lab sample gives a definitive fuel percentage.
- Do not run hard while diluted — viscosity is down and the bearings suffer.
Source found — 2026-08-05, rocker cover off
Fuel leaking from the rearmost (No.4) injector, at the top, apparently from the nut itself rather than the leak-off pipe. This is the leak-off / cap-nut sealing area, not the brazed pipework.
Consistent with every measurement: dumps directly into the rocker box then the sump, only while running, and the leak-off circuit carries ample volume to account for ~1.0 L/engine-hour.
Do not simply retorque
- Sealing washers — these injectors seal at the cap / leak-off union with copper washers. A missing, flattened or reused washer will not seal at any torque, and overtightening a cap nut risks distorting the spring/nozzle stack or cracking the cap. Renew copper washers, never reuse; torque to the Lehman spec.
- Check the leak-off return is clear — if restricted, leak-off pressure rises and forces fuel past the weakest seal. Would explain both the progressive pattern and why the rearmost injector failed first, since it typically sits at the dead end of the gallery.
Also check the cap nuts on the other three while the cover is off, and consider pop-testing No.4 or the full set — 7,000+ hours and it has been leaking.
Return-line test — 2026-08-05, diagnosis confirmed
Two observations close this out:
- Return line disconnected and blown through — completely clear, no restriction.
- Engine run with the return disconnected — no fuel came out of the engine return line at all.
The leak-off circuit should deliver a steady modest flow back to tank while running. It delivered nothing. Combined with visible fuel at the No.4 top nut and the measured ~1.0 L/engine-hour crankcase gain: the failed seal at No.4 was discharging essentially the entire leak-off flow into the rocker box and down to the sump. Large enough to relieve the whole circuit, which is why nothing reached the far end.
Back-pressure theory is dead — the line blew through clear, and there is no flow to generate back-pressure in the first place. No second fault at the tank fitting to chase. One fault: the No.4 injector leak-off / cap seal.
Repair and confirmation
Minimum fix would have been to renew the copper sealing washer(s) at the No.4 union and torque to spec, inspecting the banjo / union for cracks, and renewing the other three washers as same-age parts. Superseded 2026-08-06 by a full set of new injectors (see Replace Fuel Injectors).
Confirmation, fastest first:
- Fast — run a couple of minutes with the return routed into a jug. Fuel returning to tank again = circuit restored.
- Slower — run ~1 h and confirm the oil level holds. At ~1.0 L/engine-hour any residual leak is unmistakable.
- Replace Fuel Lift Pump Maintenance
Replaced the fuel lift pump while investigating persistent diesel dilution of the engine oil (2026-07-11 and 2026-08-03 changes, a 1 qt top-up 2026-07-29, then a severe ~5.7 L event on 2026-08-05).
Old pump inspected after removal — no fault found
Examined the removed pump: no visible damage, and definitively not a ruptured diaphragm. Possible slight gasket misalignment, but nothing conclusive.
Treat this replacement as a candidate eliminated-ish rather than a fix — the leading hypothesis is now weak.
Caveat
A diaphragm can pass fuel without visible tearing:
- a pinhole, or perished / porous material
- fuel bypassing the clamped edge if the body halves or gasket were not sealing — which matches the misalignment noticed
Also worth checking whether this pump body has a weep / tell-tale hole. That hole exists precisely so a failing diaphragm leaks externally where it is visible; if it were blocked, the same leak would route internally to the crankcase instead.
So not fully excluded — just no longer well supported.
Old pump retained for further inspection if dilution recurs.
- Change Engine Oil Maintenance
Engine oil changed — oil only, filter NOT changed (filter left in service from the 2026-07-11 change).
Condition
Drained oil again heavily diluted with diesel. Second consecutive change showing dilution — see the 2026-07-11 log and the 2026-07-29 1 qt top-up. Root cause still under investigation; suspect an injector / injection-pump leak past seals into the crankcase.
Hours
~131.5 engine-hrs since the 2026-07-11 change.
Derived from the taku RPM ledger including 59.1 h of sessions still queued on inuc1 — cloud upload was blocked by
/etc/taku/publish.tokenpermissions — so this figure leadsVesselTelemetry.engineHoursTotaluntil that backlog uploads. - Engine Oil Top-Up Maintenance
Added 1 qt (0.95 L) engine oil. Top-up between changes; see 2026-07-11 oil change note re: suspected diesel dilution.
- Change Engine Oil Maintenance
Engine oil change. Drained noticeably more oil than was added at the previous fill — suspected diesel dilution of the crankcase (fuel leak past injectors/injection pump) or excessive idling. Watch oil level/smell for diesel.
- Change Fuel Pump Oil Maintenance
Fuel injector/injection-pump oil change. ~68 engine-hrs since previous change (2026-05-22) per journey log.
- Check Steering Fluid Maintenance
Steering fluid checked. Logged retrospectively on 2026-07-29 ("32 days ago"); engine hours derived from the ledger as of that date.
- Change Fuel Pump Oil Maintenance
- Change Engine Oil Maintenance
- Replace Heat Exchanger Pencil Zinc Maintenance
- Change Rudder Zincs Maintenance
Rudder zincs changed. Logged retrospectively on 2026-07-29 ("end of April this year"); date taken as 30 April, which is immaterial for a seasonal task.
- Change Fuel Pump Oil Maintenance
- Transmission Oil Maintenance
Transmission oil changed at the start of the 2025 season, before departing Langley WA. Logged retrospectively on 2026-07-29; date approximate (immaterial for a seasonal task). Engine hours unknown — predates the 2025-07-07 hour-meter anchor.