I bolted awake at 1:15am to a shrieking alarm. We were 50 miles south of the Grand Banks, in large seas, on passage from Newport, RI to Kinsale, Ireland. I ran upstairs to the pilot house and Jennifer, at the helm, just said “high bilge water.” Yuck. Better than fire but far from good news. I ran back down stairs to get to the engine room and, yes, conditions there certainly do warrant a high bilge water alarm. I hadn’t even stopped to put clothes on yet but in the short time between the alarm firing and me arriving in the engine room, the water had come up above the bilge and the port side engine pan that forms the walkway around the engine was already awash.
We have never really enjoyed crossing oceans, but neither do we hate them or find them scary. We mostly do it “to get to the other side” and to enjoy cruising different parts of the world, but never for the enjoyment of being as sea. Generally, crossing oceans is just work and rough conditions can be tiring but usually nothing more. This is the first time we’ve felt serious doubts and even felt a touch of fear, and considered turning back. The volume of water entering the boat was simply staggering. It’s amazing how alone you can feel when looking at the engine room floor awash, the water level climbing fast, while hundreds of miles from shore in difficult sea conditions.
We expected winds in the 20 kts ranges but we were seeing a steady 30 to 35 knots with the worst gust to 47 knots. This low was worse than predicted and sea state was unusually poor. The boat was just flying around as these large and very short period waves roll past. The boat was rolling over 20 degrees and sometimes got up over 25 degrees even with active stabilization. Pitch was ranging between 12 and 15 degrees and it is, by far, the worst of the two.
Still not dressed, I went back to the engine room looking for the leak. I couldn’t find any fitting or through-hull leaking in the engine room so I continued to search the lazarette. Water was pumping down the 2” Glendinning shore power cord retractor pipe at the aft starboard corner of the lazarette. As I watched I could actually see what appeared to be waves, where the flow was steady, but every few seconds, a massive amount gushed in with wave action or perhaps boat motion. It’s amazing how much water was getting in.
We turned the boat 180 degrees in an effort to reduce the waves boarding the cockpit. Surprisingly, it didn’t seem to help much, but we left it that way for about thirty minutes. I threw on some clothes and a life jacket and worked my way carefully outside to the back of the cockpit to find the leak at the starboard rear cockpit locker with Jennifer keeping an eye on me from the salon. It was dark, cold, and when standing outside at water line level, the waves towered above the pilot house. From outside, I suspect the waves looked much bigger than they actually were. I was standing in around 6” of water but, periodically, waves rolled slowly over the transom soaking me and filling the cockpit to more than 12” of water. With a deck fuel bladder strapped down on the floor of the cockpit, I could only open the locker door three to four inches. Even that small opening is plenty to be able to see the water level at least 6” inside the locker and roughly equal to the water height in the cockpit where I’m standing.
The cockpit has seven scuppers to dump water out of the cockpit and walkway but the waves were rolling over the transom few minutes. Each brought in hundreds of gallons of water and soaked me while I worked at the transom trying to understand how the water is getting in. The deck drains become geysers when the waves hit so, rather than serving as drains, they are acting as cockpit fillers. The scuppers are designed to let water out efficiently and to slam shut when the water is higher on the other side. They do this reasonably well but water still sprays in the 7 scuppers and there is always a lot of water on the cockpit floor when we are in rough seas. As long as the water doesn’t find a way into the boat, having water in the cockpit is not an issue. I’ve seen the cockpit filled to the top and, the more full it is, the faster it dewaters. From my perspective, it’s a perfectly safe design. What doesn’t work well is the cockpit storage lockers have drains which, like all drains in rough seas, both allows water in and, if the water ever gets down below the height of the drain, it can let water out. This design just about guarantees the cabinet will be full of water when operating in rough conditions. I would prefer it stay dry but it’s not a disaster to have water in the storage locker unless it starts flowing into the boat below.
The cockpit lights were on, I was wearing a headlamp, and had a flash light so light conditions were good. But, even with good light, the entire situation just feels more difficult at night. Every 10th wave or so came over the transom and I’m soaked in cold North Atlantic water. I can see the water is going down the standpipe that feeds the shore power cord below. What prevents water ingress on this path is the collar around the hole. Until the water gets above the collar, nothing goes down into the lazarette. The problem is the water line in the locker is frequently above the height of the collar so it’s just pouring into the boat below. In fact, in these conditions, water is just about always above the collar and it’s an eye opener how much water can flow down that pipe. I understand the design point but, to work, the cupboard needs to keep water out and the header pipe inside the locker probably should be higher.
Thinking through the possible options for water getting into the storage locker the options that seemed possible were: 1) Glendinning power cord entry, 2) the right side grab rails, 3) the swim step attachment, 4) the storage cabinet drain hole, 5) the locker door, or the 6) the louvers in the locker door. Hanging on as securely as I could, I looked out over the transom to the swim step using a head lamp and a bright flashlight. It looked solid. Waves are still coming over the transom so a cold flood was hitting my chest every ½ minute or so. I checked the Glendinning power cord external cover and it’s also screwed on securely. Everything outside of the cockpit looks great.
The storage cabinet drain hole is clearly a big part of the problem and it might be the entire problem. Jennifer went to get a 1” rubber plug from our spares to seal off the locker drain hole. I installed it and then we checked the rate of water leaking down below. It’s definitely a lot better but the water is still coming in pretty fast. Probably ½ of what saw before but still faster than the main bilge pump can keep up.
Every five to seven minutes, we need to get the emergency hydraulic pump back on. Ironically, while the main bilge pump can’t keep up, the emergency pump is so incredibly fast that it evacuates the bilge in under 15 seconds. Because this pump will lose prime if the water is completely evacuated and the pump will fail quickly if run dry, it ends up being a bit difficult to deal with. To operate the pump, I have to be down below ensuring it has prime while Jen is in the pilot house turning the pump of and on. Needing us both to be available to use the pump isn’t ideal, so I will put a switch near the pump in the engine room so this can become a one person job.
Between the difficult sea conditions tossing the boat around and having to go back to emergency bilge pumping frequently, all the work to investigate the leak took longer than it should. Water flowing into the boat at 3 to 5 gallons per minute is actually a bit scary. It was far less with the plug in, but we’d not feel good until the flow rate was way below the main bilge pump capacity and preferably stopped entirely.
I realized it’s partly I’m tired, partly working at night is a bit harder, partly being completely soaked cold water, and partly I realize some of the disorientation is sea sickness. Jennifer helped apply a Scopolamine patch to combat sea sickness. In my life I’ve only had two occurrences of sea sickness. The last time was in the Gulf of Alaska in 40 kt winds. This isn’t quite that bad, but the pitching is over 15 degrees and the roll is up over 20 degrees. Sea conditions slow everything and even slight sea sickness seems to suck energy. Looking at my watch it’s just past 3am and we have now been at this for nearly 2 hours.
I went back up outside into the cockpit to understand why so much water is being forced down the Glendinning storage pipe. Having water in the cockpit is fairly normal for us in rougher seas. The leak from the cockpit boarding locker down into the laz is something we have seen before in rough conditions but only small amounts of water. It’s done that since the boat was new. Whenever the cockpit water gets much deeper than 6 to 8 inches of water we get a slight water leak. I have always found it annoying but I can’t see an easy solution and, in the past, very little water actually gets in. These conditions just seem perfect for filling the cockpit so we are suddenly getting massive amounts of water even though nothing has changed. It was clear that we needed to stop up the pipe the Glendinning shore power cord retractor uses to stow the 100’ of shore power cord below in the laz. Further complicating the investigation, the fuel bladders on deck only allows the locker door to be open around 4” so I couldn’t get things in the storage cabinet out of the way nor reach this hole that needs to be plugged.
The best approach to mitigate the inflow was to seal off the shore power cord to the standpipe and just ignore the fact that the locker is full of water. If it’s not getting into the boat in a big way, it’s not a problem. Since I couldn’t get the locker door open far enough with the fuel bladder in the way, I was not able to plug the pipe at the top. I decided to deal with the leak down below.
We still needed to stop work every 5 to 7 minutes to use the emergency bilge pump to get the bilge water back down to safe levels. It was distracting to have to stop working on mitigating the leak and run the emergency bilge pump and it was making everything take longer.
I jammed foam insulation rubber up into the Glendinning power cord entrance hole but the water pressure from the several inches of water above is sufficient that my foam blocks weren’t really have a dramatic positive effect. I then used a screw driver to force rags up into the hole to fill small gaps and this began to show some promise. Finally, I wrapped a towel around the entire assembly and compressed it on tightly using rope and a couple of heavy duty wire ties. This is essentially a marine version of what a trauma doctor would do to stem arterial blood flow to stabilize the patient to allow time for a proper solution.
The inflow was now substantially improved, but it had been nearly 4 hours and it’s both distracting and nerve wracking to have keep stopping to operate the emergency bilge pump. The water ingress rate was now only what can soak through the towels – fairly minimal.
When 100s of miles from shore or safety, this still looked to me like a lot of water but it’s actually not that much. The main bilge pump should be easily able to handle that level of flow. We now only need the emergency bilge pump every 15 minutes or so. Something must also be wrong with the main bilge pump.
Things were closer to under-control at this point so I took a short break to think things through a bit more. The water wasn’t flowing in fast at all and yet the main pump can’t catch up. Of course! The main bilge pump strainer must be plugged. Massive inflows can free up boat build debris and plug pump strainers. I was pretty confident that was most probably the problem. I went down the engine room and cleaned out the strainer. It had caught a lot of debris. The strainer wasn’t fully plugged but there was enough build debris mixed with blond hair and black cat fur to perhaps slightly reduce pump output. I cleaned the strainer and restarted the pump. It worked and it quickly filled up the strainer but we were still falling behind and needed to stop a few times per hour to run the emergency bilge pump. Something was wrong. The main pump should be able to easily manage this flow.
I decided to test the bilge pump with the thin mesh strainer removed entirely. After watching bilge levels for ten to fifteen minutes, it was clear that the main bilge pump was still falling behind even with the current small water inflow rate.
I took another break and Jen and I talked it over. Something must be restricting the output of the main bilge pump. We know that strainer is clear and we can see the pump drawing water so, for sure, it is working. This pump may not have been able to keep up with original flow but, at this point the rate of ingress isn’t that large and just about any cheap pump should be able to handle what we have coming in. We decided to change the pump.
Changing the pump involves moving six five gallon pails of oil, our waste oil container, and assorted other containers and spares. We need to get all the oil out of the way to work but need to have the heavier items secured. In these conditions, these oil containers are like small 35 lb missiles. We played with putting them in different places and tying them down but just moving them to directly behind the engine had them stable enough that the rough seas weren’t launching them and seemed safe to work.
Once the oil containers and spares were stable, I lifted out the floor boards to expose the main bilge pump. I could see I needed a 7/16” ¼” drive socket but the tools were in the laz and much of the stores that needed to be removed to get access to the bilge pump were now in the laz. So, I climbed over top all of the spares to get to then back again. I then shut off the main pump and took it apart. With it no longer pumping, we need to stop more frequency to run the emergency bilge pump. And, even though know the emergency bilge pump hardly needs to run at all, it just feels “wrong” to have the main bilge pump apart while taking on water. So I was working as fast as I could.
With the pump taken apart, I could easily see that the pump outlet check valve had corroded sufficiently that the rivet holding the valve in place has failed. Likely this happened about 5 minutes prior to the bilge high water alarm going off since it was previously cycling frequently but more than keeping up. Jennifer went to get 2 pump valves from our spares inventory. Our plan is to replace them both.
The Jabsco 34600-0010 is a nice pump and moves good volume, but these valves are one of the two weak points of this otherwise excellent pump. The first of the two design weak points is corrosion-related valve failures as we saw here and the other being the Bakelite pump base can crack if over tightened or if tightened unevenly. The second is easy to avoid but the first can fail at any time. Testing the pump is a good practice but even doing that diligently, there is still no assurance it’ll be there when you need it – it could fail 5 minutes after the last successful test. To be on the safe side, I think the valves just need to be replaced every year or, perhaps on the outside, every two. Not using corrosion-resistant rivets takes a good pump down to less-than-acceptable quality when it’s actually pretty good by most other dimensions. (Update 2017.06.10: After considering this situation carefully and reading numerous comments on it from others, we have concluded this pump has sufficient quality and volume problems that it should not be counted on for more than routine bilge dewatering).
Technically we have a high water bilge pump backing up the primary bilge pump, but the high water bilge pump doesn’t even turn on until the main bilge is nearly full. It’s tough to just look at that much water in the bilge and feel like things are under control. We might not have needed the emergency hydraulic bilge pump. It’s possible the high water bilge pump could have controlled the flow and, for sure, it could control the flow after we had reduced the rate of water ingress. We’ll just need to test it out in a safe location and see the actual real world output volume for this pump. We’ll also want to do the calculations and convince ourselves that its’s safe to allow the water levels to get up to the height of that pump in rough seas.
I put the main pump back together and started it up. Man, it was nice to hear the pump start up and then take on load as the water was pulled into the pump. The weird thing is it didn’t solve the problem. In fact, it actually seemed that it might be making the problem worse. Unbelievable! I had somehow managed to put the valves in backwards. This is an unexplainable rookie mistake but it’s been more than 5 hours working on the problem and I had only had two hours of sleep in the last 36 and I guess I just screwed up.
We again used the emergency pump to clear the bilge yet again and I went back to putting the valves into the main pump correctly. I was annoyed with myself because this is a simple job and it’s weird that I made that kind of mistake. The pump was now close to back together the second time and I was moving along quickly since I really hate not having that pump operating when we are taking on water. Wouldn’t you know it, that’s when I dropped one of the four pump bolts. Really? Even more annoying, the boat motion is such that it rolled to an impossible to reach location.
We wasted a ton of time trying to get to that dropped bolt but there was just no way to reach it. We then wasted tons of time searching through our bolts for one that would fit. I really didn’t want to run the pump with only three bolts since uneven load on the fragile Bakelite case will almost certainly crack it. Reluctantly we decided to dig out the spare bilge pump. It’s in the most difficult to access spares storage location on Dirona. Making it considerably more difficult to get to the spare pump, we were still swinging side to side 20 degrees and pitching 15 degrees. Even easy jobs take longer than you might expect.
We got the bolt from the spare pump, I installed it, and put the pump back on line and seconds later the pump was humming away. It caught up with the leak in 5 minutes or so. I never would have thought the seeing the bilge pump light going off would be reason to celebrate but, wow, it sure was nice to see the main bilge pump catch up. We now feel like we are out of the danger zone but it’s been 7 hours.
The emergency pump was easily able to handle flow and could have easily handled as much as 10x more flow. We also have an additional high-volume gas-driven backup pump that we never even needed. Since we came nowhere near to using our full onboard pumping capacity, it really shouldn’t have been that big a deal. However, the last 7 hours felt far from relaxing.
At this point, we were basically done, so I just put everything back together and secured the boat again. I don’t remember exactly when, but while working to slow the flow of water into the boat, I noticed the main fuel filter was running excess vacuum. Likely this was caused by the seas being so rough and debris in the bottom of the tank getting stirred up. Since the rush is over, I now went back down to change the fuel filter since we always want one ready to go and, since the last one plugged in less than 150 hours, we want to be ready for another quick change if we need it. Fuel filter changes are easy even in rough water so that was done quickly.
We’re both pretty tired and Jennifer is particularly behind on sleep so she went below for a couple of hours while I watched the helm. I was able to sleep for a couple of hours after Jennifer came back on the helm and we were then back on our normal watch schedule. Conditions were better than they were earlier but, even if conditions were identical, the boat handles them well and we have seen similar before. What really made this one different is we had a mechanical problem that needed immediate solution. We’ve never had that happen before in nearly 16 years of boating and nearly 13,000 hours underway.
Rough water is normally not a big deal – we actually don’t see much of it and, when we do, we are just doubly careful not to slip and fall. Otherwise, it’s not really a problem. Dealing with a serious water leak transforms what should only have been a bit of rough water to a much more dangerous situation. At sea it’s surprisingly easy to get nervous, stop thinking as clearly, and I found it was super easy to make small mistakes that cost precious time.
In thinking through why this water ingress problem was so bad, part of the problem is the fuel bladder has a slight diagonal twist to it on the cockpit floor. The two corners close off a small area in the aft stbd cockpit corner. This effectively dams a small amount of water in the after corner of the boat. There is a deck drain that that will allow water to run out but, whenever we get hit with big waves, water geysers up from that deck drain into that damned off area. When waves roll over the transom, almost the entire volume falls into this area with only a single scupper and deck drain. It’s not much water but it’s a small area so there is 12 to 18” of water just about constantly in these conditions. I suspect with bladders or without, these conditions would definitely be bring water into boat but the bladder placement might have made it slightly worse.
I decided to pump that single fuel bladder contents down below deck. We normally only pump the two bladders at the same time to keep boat trim level. But, I remember from filling them one bladder at a time, there wasn’t much of a list so we decided to pump that single fuel bladder below. It worked fine, there was only a degree and a half of list so that’s fine as well. I would prefer to pump off both tanks at the same time but don’t yet have room in the below deck tanks to be able to fit the contents of both fuel bladders below.
At this point the water ingress was almost completely stopped. Since new, if the cockpit fills with water, the storage cabinets in both sides bring in some water as does the cockpit shower at the hose area. It’s unusual to see the cockpit filled to the top but, when it does happen, the design of both cockpit lockers will always allow water into the laz. As soon as the water level in the cockpit is above 8 to 12”, there will always be some water brought into the laz.
When water inside the cabinet gets higher than the outer Glendinning pipe, it will flow unobstructed below. It’s rare, does no damage, and the main bilge pumps the water quickly. I still hate the fact that it does that and always have but it’s not that common to have massive amounts of water in the cockpit. What was different here is sea conditions were dumping water into the cockpit in higher than usual quantities. And normally, any water in the cockpit is spread all over the entire surface evenly so a given volume produces a lower average level. The fuel bladders take up ½ the volume of the cockpit so it takes less water to get a given cockpit depth. It appears to be that with the fuel bladders on deck a nuisance leak becomes a much larger problem under some wave conditions and boat directions.
We’ll give some thought on how to mitigate this issue. But, had the main bilge pump valve not failed during the night, I wouldn’t have been up at 1:15am and this would have been nothing more than the bilge pump coming on periodically. Clearly you don’t want to see frequent bilge pump cycling, but it’s not an emergency situation. The biggest problem here appears to be that an inch or two of water in that locker flows below. And it’s almost impossible to prevent there being an inch or two of water in that locker. A good solution would be to have the Glendinning feeding through a pipe that would allow the water to get as deep as eight or ten inches without bringing any water in, instead of the current pipe that is less than an inch high.
I’ll need to come up with a better solution to seal off this large water ingress path. My first choice is getting the cabinet sealed well enough that it can have water above it and still not leak. That would be asking a lot from a weather strip so I’m skeptical that will work. Failing that, I’ll need to find some way to close off the Glendinning hose retraction opening down to the storage area below in the laz.
We have lots of backup pump capacity (see Fighting Water Ingress), but generally I don’t like any water inside the boat and never want the main pump on even once in a shift. Normally the bilge pump on Dirona only turns on periodically and briefly in rough seas and it often goes many months without cycling at all.
The weather models indicated that the sea conditions coming up in a couple of day after this event would actually be rougher. I didn’t expect it would be the slightest concern. As long as water isn’t getting into the boat, expected conditions will rough but perfectly safe and hardly worthy of mention. Just like most of our ocean crossings.
In spending a half day fighting the water ingress, we learned several important things about our bilge dewatering systems. The first is the primary bilge pump is not that reliable and doesn’t pump that much volume. The second lesson is the high-water bilge pump, although it produces excellent volume, is mounted way too high in the bilge to serve as a primary high-volume dewatering system. The third lesson is a confirmation that the hydraulic bilge pump moves absolutely prodigious amounts of water, but ends up being a two-person job to operate. It pumps so fast it can easily run dry and/or lose prime so someone needs to be down in the engine room to monitor it while a second person operates the off/on switch in the pilot house.
These are the changes we plan:
- Don’t count on the low-volume and low-reliability Jabsco 34600-0010 for anything more than bilge drying.
- Install a Rule 4000 pump just above the main bilge-drying pump. This will be the new high-water bilge pump (primary bilge dewatering system). We will leave the Rule 3700 in place as the backup high-water bilge pump.
- Install an off/on switch for the hydraulic bilge pump in the engine room at the pump to allow single-person operation. We’ll leave the pilot house switch in place as well.
- Install a third high-water bilge alarm. We had no faults, but this experience underlined that catching these problems early could be the difference between success and failure so we want triple redundancy.