Thursday, November 10, 2016

Broken Turbo Studs - One Last Time

Back in this article I mentioned that on our way north last spring we had another occurrence of broken turbo studs.  It had happened about a year before up in Juneau, and we thought we had the problem licked, but apparently not.

When the studs crack, the seal between the exhaust manifold and turbo is compromised, and because both the manifold and turbo are cooled by engine coolant, it all starts leaking out.  Sooner or later this completely disables the engine because without enough coolant, it will overheat.  Fortunately both times it happened we were on our way to port, and made it in without overheating.

Below you can see the exterior evidence of the failure.  The yellow stuff is coolant.  You can also see the vertical brace installed last year to help take the weight of the exhaust off the turbo.   It may have helped, but clearly wasn't a complete solution.

Leaking coolant due to turbo stud failures

In the picture below, once the turbo has been removed you can see the two upper studs have parted.  The one on the left broke leaving a good stub to grab onto to remove the stud.  But the one on the right broke off about 1/2 thread inside the manifold, so I had to drill it and use a bolt extractor to get it out.  You can also see the coolant passages.  The center opening is the exhaust passage, and the oblong openings above and below are the coolant passages.

Cracked turbo mounting studs

Since the exact same thing happened again despite installation of a support brace after the first occurrence, further investigation was clearly warranted.  One common factor is that leading up to both failures, we were running the boat pretty hard - on the order of 2000 RPM out of a max of 2200.  This is absolutely fine - the engine is rated for 24x7x365 operation at 2100 RPM, and up to full 2200 RPM 16 out of every 24 hrs. - but it's different from our normal cruising at 1600-1900 RPM.

To make a long story short, the picture below will show what's happening.   The tall structure circled in red is rigidly attached to the turbo.  Because it's rigidly attached, when the engine vibrates, the exhaust structure has to vibrate with it.  But the structure has a very long lever arm, so requires lots of force to make it move along with the engine.  It's like holding a baseball bat and trying to shake it around.  If you hold it in the middle where it's center of mass is located, it's not real hard to shake it back and forth.  But if you hold it from the handle end, now the center of mass is out maybe 2' from your hand and it's much harder to shake it back and forth.  That's what's happening with this exhaust structure, except instead of weighing 2 lbs like a baseball bat, this structure weights over 30 lbs.  Trying to move it in time with the engine vibrations simply overwhelms the turbo mount and breaks the studs.

Excessively tall and heavy exhaust riser

After talking with a number of people, I decided to pursue a two part solution.   The first, and most important part, was to install flex piping in the vertical exhaust section.  Doing so would reduce the mass directly attached to the turbo, and decouple any engine vibrations from the rest of the exhaust structure.  The second part was to add a new set of spring hangers over the big flange to carry the weight of the exhaust over the turbo.  Ideally the exhaust structure should mate with the turbo elbow, but exert no pressure on it in any direction.  It should just float above it.

I probably could have ordered parts and gotten all this done in Port McNeill, but we didn't want to lose any more time than necessary, so I decided to haul everything down to Seattle in a rental car, get the work done there where I knew the people and places, then come back and install it.  So I removed all the insulation blankets, that exhaust riser, and the turbo.  Got a rental car, and drove for 10 hrs including a ferry crossing and boarder crossing to get back to Seattle.  I had lined up Hatton Marine to do the exhaust modification, and they started working on it before I even left Port McNeill.  They also pulled and ordered all the parts that I would need for reassembly.  And I lined up Ballard Insulation to modify and/or make new insulation blankets as needed to fit around the modified exhaust, and around the new mounting brackets for the hangers.

The next morning I was at Hatton at 7:00, dropped off the exhaust structure, picked up the parts, then went to visit some friends while I waited.  Hatton finished early in the afternoon, and I hauled everything over to Ballard Insulation for them to work on.  By 3:00, everything was done and I was back in the car on my way north with plans to stay the night in Nanaimo, and continue on to Port McNeill in the AM.

Below is the modified exhaust riser with attached mounting brackets to connect to the hangers.  And note the buckets of coolant.  It's about 12-13 gallons that has to be drained out to do this job.

Modified exhaust with flex section and new hanger brackets

After arriving back in Port McNeill, I spent the next two days putting everything back together again.  First, the turbo goes back on along with the oil deliver and return pipes, the compressed air pipe, coolant return, and the vent pipe.  Oh, and don't forget new mounting studs and nuts.

Turbo reinstalled

Then the turbo elbow gets installed.  It's easier to attach the elbow to the turbo and ensure a good fit, then align the rest of the exhaust to just float over and kiss the square flange that you see below.

Exhaust elbow attached to turbo

And fast forward to installed hangers and fitted exhaust section.  Once everything is in place, the spring hangers can be adjusted and the various joints rotated to get everything perfectly aligned, and with no weight on the turbo flange.

Modified exhaust installed, with flex section and spring hangers

Here's a video overview of the modification, also showing how the mating flange should float freely over the turbo elbow.

One obvious question is why, after building 60 of these boats, am I now having this failure?  I think the two pictures below say it all.  The first picture is of a friends earlier vintage N60.  His boat, and pretty much the entire N55/N60 fleet were built with the Deere 6081 engine.  The engine and turbo stand taller, with a resulting shorter exhaust riser.

Compare that to the next picture of my boat with the newer Deere 6090, introduced to meet EPA Tier II regulations.  The engine sits noticeably lower, resulting in a significantly taller exhaust riser.  This larger structure apparently is just enough to overwhelm the turbo, where the older arrangement was fine.

Shorter riser on Deere 6081 engine

Taller riser on Deere 6090 engine

So that was our major repair last spring.  But to be sure you don't go away thinking that I do nothing other than fix things on this boat, I have come up with a new benchmark for measuring a boat's condition.  You judge based on how quickly you run to the hardware store or marine supply store when you arrive in port.  The faster you run, and the more you buy, the more trouble you are having with your boat.

I'm very pleased to report that after this repair, and for the rest of the summer, I did not make a single purchase at a hardware store or marine supply store.  Based on that, I'd say the boat is running pretty darn well.

Sunday, November 6, 2016

Alternator Reconfiguration for Improved Power

Like many boats, our main engine is equipped with two alternators.  One smaller 85A unit charges the starter battery, and one larger 190A unit charges the house battery.  Ours also came equipped with Balmar smart regulators controlling each alternator to provide 3-stage battery charging for each battery bank.  This is a good setup, and worked fine, but for a handful of reasons I wanted to make some changes.

Dual alternators

Dual Balmar regulators

The first motivation to make changes was because the 85A start battery alternator was getting very little use.  Under normal conditions, an engine start battery doesn't get used very much.  When you run the starter, it draws a huge amount of current, but the engine typically starts within a few seconds, so the total amount of power draw out of the battery is very small.  As a result, it takes very little time to recharge it, after which the alternator is just spinning and producing little to no power.  That's 85A of potential charge power sitting idle, and meanwhile the house alternator is running full tilt for hours on end recharging the house battery after a day and night at anchor.  So I wanted to use that alternator to help re-charge my house battery faster while under way.

The second motivation is that I also had plans to add 240V inverter service on the boat so we could do laundry while under way.  This will be the subject of another article.  Up until now we had to run the generator, which works, but with only the washer and drier running, it was a very light load for our 20kw generator.  So running off the main engine alternator(s) via an inverter makes sense in the long run.

All this led to the same conclusion; combine both alternators and direct them to the house battery, and charge the start battery some other way.  The 190A alternator puts out a total of about 5kw, and the 85A unit puts out about 2.2kw.  By combining them it's almost a 50% increase in charging capacity, and is a close match to my 7kw of inverter capacity.

Problem 1: How to charge the start battery?

If I take away the dedicated alternator charging the start battery and use it to charge the house battery, it's just a matter of time before the start battery would go dead.  So I needed to come up with a new way to charge the start battery.  The good news is that there are lots of ways to so it.  The bad news is that you need to pick one and live with the trade-offs.  If boats teach you nothing else, they do teach you about trade-offs.  Here are the different approaches and the associated trade-offs as I see them.

Options A: Parallel the start and house batteries while charging the house battery.

There are a number of ways to do this, but they all have the same effect.  When the house battery is getting charged, simply connect the start battery to the house battery and let it go along for the ride.

- It's a nice simple approach
- There are a multitude of devices to automate the operations
- As long as the house and start battery are the same type, i.e. AGM, flooded acid, Gel Cells, the house and start battery will share the charge pretty well.
- When on a shore charger, both batteries are maintained.  This is important when leaving the boat for an extended time to compensate for self-discharge in the battery.

- Doesn't work if you have different types of battery for house and start, e.g. flooded for one, and AGM for the other.  They won't share the charge properly.
- Even though the battery will share the charge, the charge cycle will be tailored to the house bank, not the start bank, and may not be all that good for the start battery.  This is probably the biggest issue, and is most prominent with 3-stage chargers.  Here's a good example of how it can be a problem.  Say you are cruising, anchoring every night, and running for 4-6 hours each day.  While at anchor your house battery gets drawn down, so the next day when you are underway your charging system will be operating at higher voltages for most of the cruise as it works to recharge your house battery.  Meanwhile, your start battery only got a slight discharge when you started your engine, but because it's along for the ride as the house battery gets charged, it's going to be charged at a higher voltage all day when all it really needs is a relatively short recharge.  As a result, your start battery ends up getting over charged.  It won't be sudden death, but it likely won't last as long as it would if it were charged according to its needs rather than the house battery's needs.

Lot's of devices are available to do this battery paralleling.  Some support very high charge current, and some support lower current, but they all have the effect of sending the start battery along for the ride as the house battery gets charged according to its needs.

Example products:
Xantrex Echo ChargeBlue Seas ACRs

Option B: Use a Battery Charger

The other approach is to connect a dedicated battery charger to the start battery.  Ideally it would be DC powered by the house battery, and automatically switch on only when the house battery is being charged.  If your house battery isn't being charged, you probably don't want the start battery charger to drain it down.

- The start battery gets charged according to its needs rather than the needs of the house bank.  This gives max life to the start battery.
- The only power consumed is what's needed to recharge the start battery.
- Start battery can be maintained at full charge while on shore power for extended periods of time.

- Dedicated charger are generally more expensive than battery paralleling devices.

Several devices are available, including:

Balmar Duo Charge
Mastervolt  MAC/Magic

I elected to use a battery charger to ensure that our start battery received a charge appropriate to its needs.  Our cruising often involves the exact scenario that leads to over charging if you use some sort of combiner, so this seemed to make sense.

I'm generally not a fan of Mastervolt.  Their products are pretty good, but poorly documented, ignore the needs of North American split phase power (only really applies to their inverters), and have little to no tech support in the US.  But in this case I picked them, and it's entirely because of what I consider to be a major flaw in the Balmar Duo Charger.

The Duo Charger is limited to 30A, which is fine.  But if the battery calls for more than 30A - say because you had to crank your engine a whole lot to get it started after changing the fuel filters - rather than limiting charge current to 30A and running for however long it takes to recharge the battery, the Duo Charger shuts down.  It periodically starts up again and retries, but unless the Battery Charging Fairy came and recharged your battery in the mean time, there will still be a call for more than 30A and it will shut down again, over and over, and never recharge your battery.  They do have a provision for an external bypass relay.  But it's manually operated, so no better than the manual battery parallel switch that I already have.  So I consider this to be a major design flaw, and instead used the Mastervolt Magic 24/24-20 which is a 24V to 24V, 20A 3-stage charger.  It works well, recharges the start battery quickly, then maintains it at a comfortable float voltage for the rest of the time.

Matervolt Magic 24/24-20 charging start battery bank

Now, with an alternate way to charge the start battery, I was free to move on to part 2 of the problem - how to parallel the two alternators.

Problem 2: Paralleling two alternators

How you regulate multiple alternators depends greatly on how the alternators, engines, and batteries are set up.  The alternators can be on the same engine or on different engines.  And the alternators can be connected to the same battery bank, or different battery banks.  Unless they are on the same engine, and connected to the same battery bank, the alternators need to be individually regulated.  This was the case originally on my boat where the two alternators were on the same engine, but connected to different battery banks.  As a result, each had its own Balmar regulator.

But with the reconfiguration, both alternators would now be connected to the same battery bank, making it possible to regulate with a single regulator.  It's still fine to regulate the alternators individually, just not required.  Some friends on MV Dirona elected to regulate them separately as described in this article.  I decided to regulate them off of a single regulator for two simple reasons.  First, I could stash the second now-unused regulator away as a spare in case the other failed.  Second, I wouldn't have to worry about balancing the alternator outputs since they are self-balancing with a single regulator.

You might be wondering how this self-balancing works.  If not, skip ahead.  But it's really pretty simple.  The Field terminal on the alternator controls its output.  If you apply 0 volts to the Field terminal, you get zero output.  If you apply full battery voltage to the field terminal, you get full rated output.  And it's proportional in between.  The regulator controls the alternators by varying the Field voltage.  So if the regulator applies half voltage to the Field terminal, each alternator will put out half of it's rated current.  That means the big alternator will put out 95A, and the small one will put out 42A.  Each is always doing it's proportional amount of work.  It works really well, but does require that they be driven by the same engine, and be connected to the same battery.

Rewiring was actually very easy.  All I did was move the small alternator's output over to the big alternator's output, connecting them together right at the current shunt that measures the big alternator output current.  Wired this way, the pilot house meter shows the combined current of the two alternators.  I also had to up-size a couple of fuses from 200A to 300A to handle the extra charge current.  Fortunately the cables carrying all this current were already oversized (4/0), so capable of carrying the increased power with only about 0.7% voltage drop.

This is probably a good place for a little caution.  It's really important to check the entire wire run from the alternator to the battery to be sure all the intervening wires are sized to handle any increased current.  And this includes checking the ground path as well as the positive path.  Electrical fires are not a good thing.  And also recalculate the voltage drop.  3% would be the max, and even that is a bit much when you are working with charge voltages where 0.1V can be significant.

Next, I removed the now-unused second Balmar regulator and stashed that in the spare parts bin, and jumpered the field connection between the two alternators.

Another lesson in boating it to always expect the unexpected, and guess what.... This was no exception.  Initial testing was not producing the expected current out of the smaller alternator.  But why?

While poking around, I noticed something odd on the small alternator.  In the picture below, the box on the back of the alternator has two wires coming out of it connected to the positive and negative posts.  That's the way built-in fixed voltage regulators are set up.  And you see the two small holes in the box?  That's where you would reach in with a small screw driver to adjust the fixed charged voltage.  This alternator looks like it has an internal regulator rather than using an external regulator.

Alternator with unexpected internal regulator

I sent this picture to Balmar and after a bunch of head scratching they agreed that this alternator was equipped with an internal regulator, even though the alternator part number was for an externally regulated model.  So my start alternator was internally regulated, even though it was wired to an external regulator that was actually doing nothing more than turning on the internal regulator.  How messed up is that?  It's unclear where this mix-up occurred between Balmar (supplier of the alternator and regulator) and Cascade (supplier of the packaged engine and charging system), but at this point it doesn't really matter.  Fortunately Balmar was only about a 45 minute drive away, so I packed up the alternator and took it to them, and they swapped out the internal regulator box for a box wired for external regulation.

Once reinstalled, testing now showed it working exactly as expected with each alternator putting out its portion of the charge current across different loads.  And I now have a whopping 275 of charging current to get our house battery up to full charge between anchorages.  And when the charge loads are more modest, the work is shared between the two alternators rather than concentrated on one while the other loafs around doing nothing.

Wow, that ended up being a much longer article than I expected, but hopefully it will help others who are evaluating alternative charging systems.

Sunday, October 30, 2016

Kuiu Island, and more

Kuiu Island in part of the SE Alaska's Alexander Archipelago.  It's actually the 15th largest island in the US, yet only has a population of 10 people.  All of the island is in the Tongass National Forest, and parts are also within Tebenkof Bay Wilderness area, and the Kuiu Wilderness Area.  We spent a good chunk of our time in Alaska this summer circumnavigating and exploring Kuiu Island, and the neighboring areas.
First stop was kind of in the other direction, with a trip to Misty Fiord.  I never tire of all the water falls

Once in Misty Fiord, we anchored for a couple of days and were treated to a visit by a young bear.  Here he is digging for some chow with crows hoping to get some left overs.

Some other people started to approach in a dinghy, and this got the bear's attention.   They are pretty funny looking when they stand up.

Here's a another example of a water fall up close.  The colors are just amazing.

Here's Misty Fior living up to its name, rainbow and all.

Walker Cove

One stop was Coronation Island off the southern tip of Kuiu.  We were in a protected bay, but it was very windy and it was just coming from everywhere.  I have never experience such confused wind.  Here's out anchor trace for the 12 hrs or so that we were there.  You can see the boat was all over the place.
Coronation Island Anchoring

Kuiu Island is home to a LOT of otters, especially in Tebenkof Bay.  Here are three just chillin'

And here is a mom with her baby checking us out.

OK, this waterfall in Red Bluff Bay is really big, and hopefully this sequence of pictures will show it.  Below is the full water fall.   The part circled in red is the next picture.

Here we are zoomed in a bunch, but wait.  The red arrow is pointing to a red Kayak.  That's Laurie out for a paddle.

 After we got around Kuiu we stopped in Farragut bay and went to visit Farragut Farm.  They are completely isolated from everything with the only way in and out via boat, and only at high water.  And once ashore, it's still a 1/2 mile walk to get to the farm.  But despite this, they run a vibrant business selling at the Petersburg market once a week, and supplying fresh veggies to a variety of businesses and small cruise boats throughout the season.  Bo and Marja run the whole operation themselves.  We got a great tour and stocked up on veggies for the rest of our trip.

Bo from Farragut Farm

Farragut Farm

Next stop was Peterburg for the much anticipated Nordhavn Rendezvous.  It was organized by a couple of owners, and blossomed unto the largest gathering of Nordhavns ever, with over 30 boats and 100 attendees.  A good time was had by all, but we got a real curve ball the day before the event started.  Our daughter who lives in Seattle broke her leg, required surgery, and was facing 10 weeks of no driving, no carrying anything, etc..   It's amazing how much we take for granted being able to move around and do things with our hands, and equally amazing how helpless you can be without it.

So Laurie flew back to Seattle to help with everything through the surgery, and I started sorting out how to get the boat back to Seattle.  We had originally planned to keep cruising until around the 1st of August, so this cut us short a bit, but not by too much.  And besides, you need to take what life throws at you.  I'm just glad we were able to change our plans and help out.

I was able to pull together family for crew in two shifts.  First was brother-in-law Paul who flew up to Petersburg and helped get the boat down to Port McNeill in BC, Canada.  One of our stops was Prince Rupert where we checked into Canada, and also checked out the new Cowe Bay Marina.  It's good, all things considered, but Prince Rupert is a pretty exposed harbor.  We usually anchor in a cove nearby that offers better protection, but it was nice to tie up and have access to the town.  Before this new marina, there were almost no accommodations for transient boats.

Prince Rupert

Prince Rupert

We had a bunch of long days, but made good time and didn't get delayed by weather, so arrived in Port McNeill a day early giving Paul time to go on a fishing expedition.

Paul catches fish in Port McNeill

Paul and brother Dave crossed paths at the Vancouver Airport as Paul departed and Dave and significant-other Scottie joined me for the second stretch down to Seattle.  Once again we had no trouble with weather and made excellent time, arriving a day or two early, and affording Dave and Scottie some time to explore Seattle, including an obligatory visit to Bongos for Cuban food.

Brother Dave at Bongos.  "Where's my food?"

Dial forward 10 weeks, and Lindy is back on her feet, driving, rock climbing (indoors), and rapidly regaining all her strength and agility.

Wednesday, September 7, 2016

Engine Room Cooling

Keeping a boat's engine room cool, relatively speaking, is very important.  When things get too hot, all sorts of wear and deterioration starts to accelerate, including wear on you if you need to spend any time in there.  Spending any length of time in a 130 degree space is not just uncomfortable, it's down right dangerous.  But if something breaks, you gotta go in.

Speaking of 130F, that's one of the magic numbers.  A variety of equipment is specified up to a max ambient temp of 130F.  Go above that, and you are on your own.  Warranties are void, etc, etc.  I say that 130F is one of the magic numbers, because there is another one.  30F delta-t.  Delta is a Greek letter, and is used in science to represent the difference between two numbers.  So 30F delta-t means a temperature difference of 30 degrees Fahrenheit.  And the difference is between the outside air temperature and the temperature in the engine room.  If you want to run your engine room electrical equipment and machinery within the limits allowed by the manufacturers, you need to stay under BOTH numbers.

The effectiveness of an engine room's cooling system will determine the delta-t for the boat.  Cooler is always better, but this means that a cooling system must keep the ER temp at no more than 30F over the outside temp.  This in turn limits the boat to operation in 100F outside temp, because if it gets hotter than that outside, you will start to exceed the 130F absolute temp limit.  If, on the other hand, you can limit delta-t to 20F, your ER runs cooler under all circumstances making it more hospitable to you personally, and it allows for operation in a hotter ambient temp.  So the goal is to keep the delta-t as low as practical.  Everything in a boat's design is a trade-off, so these goals are often hard to meet, or in some cases not met at all.

The key to making this work is to apply a little bit of science and engineering.  You need to look at:

1) What are the sources of heat and how much do they generate.

2) How much air flow is required to remove that heat and maintain a reasonable delta-t

3) How will you get that air into and out of the ER

Sources of Heat

The biggest source of heat in your ER is your engine (or engines).  Additionally there may be a generator and some other gadgets, but the main engine will dominate, so that's the thing to focus on.

Engines work by burning fuel.  Some of that energy is turned into propulsion power. but most of it (about 70%) is waste heat.  The majority of waste heat is removed via the engine's cooling system and ends up in the surrounding sea water.  In the case of a keel cooler, it transfers to the sea water that passes by the outside of the boat.  In the case of a wet exhaust, it's transferred to the sea water that is pumped through the heat exchanger and exhaust system.

But there is still a significant amount of heat that is dissipated into the surrounding engine room.  It's only about one tenth the amount of heat that goes out through the cooling system, but it is still a lot.  One way to look at it is that you have a 1000lb to 2000lb radiator sitting in your ER running at 200F.  That warms things up.  The amount of this "ambient heat rejection" is specified by the manufacturer for every engine so whoever is designing around it can figure out the cooling.  Below is the spec page for my engine, with the heat rejection number highlighted.  The heat rejection is 22kw.  A little electric space heater is about 1kw, so it's like having 22 of them running in your ER.  It's a lot of heat to remove.

Specs for Deere 6090AFM75 engine.  Heat rejection highlighted.

Airflow Required to Remove Heat

Getting that heat out of the ER is usually accomplished by air flow, pulling fresh air in from the outside, passing it through the ER where it heats up, then dumping it back outside.  Physics is everywhere, and there is a formula relating heat, air flow, and delta-t.

CFM = (55.0 * BTU/m)/ DeltaT

CFM is cubic feet per minute of air flow.  BTU/m is British Thermal Units per minute and is a heat rate.  And Delta-t we have talked about.  So with this formula you can figure out the air flow required to maintain a certain delta-t, given the heat rejection rate of your engine.

Or, looking at it another way;

DeltaT=(55.0 * BTU/m)/CFM

This tells you what the Delta-T will be for your engine's heat rejection rate and some given air flow rate.

When you run the numbers on my engine, it says that I need about 3500 CFM of air flow to maintain 20F delta-t.  I think 20F is a good number to strive for.   Under most circumstance it will keep your ER at 90-110F which is tolerable, and it means you can run in 110F ambient if required.  Lower would be wonderful, but as you'll see in the next section, it gets harder and harder to move the requisite amount of air, so you need to find a balance.  So I set 20F as my goal.

How to Move Lots of Air

This is where things get challenging.  Up to this point it's really easy to calculate how much air flow is needed, and it's really easy to say you want more.  But now you have to find a way to actually make it happen and this was the problem on my boat.

Are moves easily in open spaces, and is progressively harder to move as space becomes more confined.  But this is often overlooked.  Take breathing as an example.  You can break in and out easily because your wind pipe and mouth are large enough.  There is no restriction.  Now try breathing through a 1" hose.  You will probably be fine, but will surely feel the restriction as you push and pull air through that hose.  Make the hose longer, and it gets harder.  Now try breathing through a straw.  Good luck with that.  It's just too small to move the requisite air.  This was a major problem on my boat.

As originally built, there were two blowers serving the ER.  One was rated at 220 CFM and connected to a 3" hose, and the other rated at 1200 CFM and connected to a 5" hose.  Each hose was then about 10' long.

When you look at the performance specs for a fan, it lists the CFM rate in free air, i.e. with no restrictions.  That's like you breathing, and is the best you will ever get out of the fan.  Then they list the performance at different levels of back pressure which is the restriction imposed by a duct or hose.  This is like breathing through various size hoses.  Once again, physics is everywhere, and there are formulas for calculating the back pressure when trying to move various quantities of air through various size hoses and ducts.  There are simple back pressure calculators on line, and I made use of this one.  When you run the numbers, the back pressure trying to move 1200 CFM through 10 of 5" duct hose is 2.5" of H2O.

Now take a look at the performance specs for that 1200 CFM fan.  In the chart below, you can see that with 2.5" H2O of back pressure, the fan is completely stalled and not moving any air at all.

Air flow curve for Dayton blower.  It stalls at a little over 2" of H2O
That's not good.  Now in practice, as the duct resistance increases, air flow decreases, and eventually the war between the duct resistance and the fan finds a point of equilibrium along the line in the chart.  As part of this project I bought an inexpensive airflow meter to help see exactly what was going on, so I measured the air flow through the 1200 CFM fan.  It was 370 CFM.  This is the big fan in the ER cooling system, and it's only moving one tenth of the required air.  The situation for the other fan was similar, with actual air flow of 70 CFM vs the free air spec of 220.  The obvious conclusion is that both fans are contributing very little to cooling the ER, and might even be a net negative given their power consumption.  Between the two fans they represent about a 1kw electric load while underway.  That power is generated by the engine alternators which dissipate more heat the more they are loaded.  The moral of the story here is don't try to breath through a straw.

Large blower sucking through a straw.  Hint - it doesn't work very well

As originally build, my boat had another fan about half way up the exhaust stack, and that's another example of what not to do.  Air, water, electricity, and probably lots of other things follow all possible paths, and do so in proportion to the resistance of each path.  A waterfall will have a main channel, and tributary paths, each sized according to how much water can squeeze through each path.  Air does the same thing.  This upper stack fan had a hose that hung down inside the stack cavity about 3 feet, then it blew out into a vented compartment on the boat deck.  I think the idea was to pull air up the stack and push it out through the compartment.  But in actuality it pulled air from down the stack, from up the stack, and from any direction it would flow.  The net result is that is too did pretty much nothing but consume electricity.

What does effectively move a lot of air through the ER is the same thing that generates all the heat - the engine.  It's actually a pretty large and very effective fan, pulling air from inside the ER, and expelling it out the exhaust.  Going back to the engine specs, it tells us that my engine moves 730 CFM of air.  So far, that's the biggest air mover on the boat.

Adding it all up, we have 730 from the engine, 370 from the big fan, and 70 from the little fan.  That's 1170 CFM total.  With that amount of air movement, the predicted delta-t is 52 degrees.  That's obviously not acceptable.  They cooling system as originally built had no chance of succeeding.

Stepping back, the problem is the size of the air passages.  We haven't talked about the intake side and that's just as important.  On my boat there are two intake vents, one on each side of the cockpit.  These lead down through a large tunnel and emerge through apertures in the ER.  Running the back pressure calculator on those passages suggest that they are large enough to carry the desired 3000 or so CFM without creating undue resistance.  So the intake side looks good.  The outflow side of the equation, however, is a big problem for those two meager houses.

The Light Bulbs Lights

After a while the light bulb finally lit, and it became evident that the exhaust stack was a good size cavity leading straight from the ER to the outside.  It's much, much bigger than the hoses.  Running the calculations using it as an air duct shows it can move 3000 CFM without excessive back pressure.  Perfect.   If a way could be devised to pull air up the stack shaft, there would be enough space to actually move the required amount of air.  With this approach, the airflow path would be in the cockpit vents, into the back corners of the ER, forward through the ER, then up the stack right over the main engine.

The best place for fans was in the upper stack at the boat deck level.  Looking at the picture below, this structure is all open inside with storage space below by the door, and open space with vents above.  Then there is a removable panel that runs the full height separating the cabinet from the stack.  So the wall with the port light in it separates the cabinet from the stack, with the removable panel inside the cabinet.

Boat deck cabinet and air vent

The first attempt at this was done in Dana Point when we were commissioning the boat and struggling with high engine room temps.  The idea was to eliminate the silly fan with snorkel hose (it previously lived in this cabinet), and install another 1200 CFM fan in the divider panel.  The fan would pull air from the stack and exhaust it in the cabinet where it could flow out the vents.

Below is the fan mounted on the divider panel, and below that you can see the panel's full height looking up inside the cabinet.  Remember this for later on...  Oh, and ignore that red hose.  It's the exhaust for the shroud blower that cools the exhaust pipe jacket.

Fan pulling directly from stack cavity, exhausting into cabinet

View of divider panel between stack and cabinet.  A perfect fan mount location

The idea was good, and seemed to get ER temps down to the magic 30F delta-t.  But that didn't last, and later I would find that on long runs the temps would still climb over 30F.

This biggest issue with this first step is that it was only a first step, and still suffered the same flaw as the snorkel hose fan.  Air follows all available paths, and this was pulling just as much cool air down the stack as it was pulling hot air up the stack.  What was still needed was some sort of baffle in the stack to seal off the area above the fan and force all air to be pulled up from the ER.  In the spring of 2015 I fabricated a simple baffle made out of galvanized duct material.  Although not a perfect seal, it had the desired effect of forcing the air flow to come up the stack.  For the first time, air flow measurements in the stack were right about 1200CFM, exactly what the specs for the fan stated.

Air baffle to force air flow up the stack from the ER, rather than down the stack

We ran for the summer of of 2015 like this, and although it was a big improvement, we were still hovering around 30-35F delta-t.  Running the calculations with the full 1200CFM of air flow, plus the engine air flow, it predicted 33F delta-t - just what we were seeing.  The problem remained a need for more air.  Something between 2000 and 2500 CFM in fans, with the rest from the engine would get it to the desired 20F delta-t.

One experiment I tried over the summer was shutting down the two engine room fans.  Together they were consuming about 1kw of power, and only moving a small amount of air.  I was curious what would happen if I just shut them off, so I tried.  The answer...nothing.  Turning them off had no impact at all on ER temps, so I left them off permanently, and when we returned to Seattle, I removed them completely and just connected the ducts through to the ER with no fans.  Now they serve as supplemental passive air intakes.

In the fall of 2015 I visited Delta-T Systems, manufacturers of fans and systems for boats, and sat down with one of their applications engineers to review my plans and calculations, and see if we could pick some fans for the job.  Long story short, he agreed with the calculations, and I selected their Small AC Axial Fans.  AC power meant no rewiring since AC fan power was already present in the cabinet.  The 11" model is quiet, consumes only 185W of power, and can move about 1000 CFM.  I bench tested one at home, and the thing took off across the bench from the air blast.  Their size allowed be to fit three of them in a vertical row in the cabinet divider panel, and still have space below in the cabinet.  One down side of the first fan installation is that it consumed the entire cabinet.

I had to fabricate a new divider panel and mount everything up, but was able to do it out on the bench.  Below are the three fans ready for final wiring and then installation as an assembly.

New triple decker fans on new divider panel

Another view of the new fans

The picture below shows the approximate locations of the stacked fans once installed in the boat deck cabinet.

X-ray vision view of fans inside cabinet

In conjunction with installing the new fans, I need to relocate the baffle.   The original baffle was just above the original fan location, and the new triple decker fans were located up towards the top of the cabinet.  So I built a new baffle which is nothing more than two pieced of fiber board that fit into this square recess where the stack transitions from the boat deck to the flybridge.  This picture shows the location, but not the baffle itself.

Looking down stack from fly bridge, and convenient inset for baffle boards

In January of 2016 I finished installing the fans, and even wired up thermostats in the ER to switch the fins on incrementally as the ER temp climbed.  But when I fired them, I was in for a great disappointment.  Instead of measuring 2500 or more CFM as expected, I measured 1300 CFM.  I was crushed and perplexed but had a bunch of other projects that needed attention, so I left it all for later investigation.  After all, it was moving more air than the previous system, and consuming a LOT less power, so it was still a step forward,  Just not as far as I had hoped.

Over this past summer (2016) we observed a pretty consistent 30F delta-T on long runs.  My suspicion was that the cabinet was imposing too much back pressure on the fans, and not allowing them to move their rated air flow.  The cabinet has a few baffles to prevent water entry, and the vent slits are potentially restrictive, so very late this summer I started to experiment on a couple of long runs where there was plenty of time for the ER to heat up.

First I removed the upper grill, and much to my pleasure saw a distinct drop in ER temp - about 5F over a couple of hours.   Then I opened the cabinet door, and the temp dropped again down to a delta-t of 18F.  Wow, all I needed to do was open that silly door.  A bunch of subsequent experimenting, including taking air flow measurements in the stack, showed that opening the door had the biggest impact, followed by opening up the upper part of the cabinet.  With everything opened up, I can get almost 2000 CFM of air flow which is much closer to the 2500 or so that I expected.  And the upper most fan is only 4-6" away from the cabinet with no relief from opening the vent grill or door, so probably explains the missing air flow.

But now I know the system works, I know it can achieve 20F delta-T, and I know what final modifications I need to make.  I have two 12" square stainless grills on order, and plan to install one in the door, and one up towards the top of the blank space directly in front of the uppermost fan.  I also ordered a 4x18" grill to install immediately above the door, but will probably wait to see how things work with the other two before I cut that one in.

Door opened and upper grill removed to reduce back pressure in cabinet
 Alternate Approaches

There are always lots of ways to skin a cat, and ER cooling is no exception.

Many people have upgraded various fans, but kept the basic cooling arrangement.  This helps a bit, but if you go back to the fundamental problem which in adequate ducting, it will never really solve the problem.

Where a number of people have found some level of success by installing large fans over the intake vents in the ER.  I believe this works because it achieves the same air flow as what I've done, just in a different way.  By pressurizing the intakes, you force air into the ER, and it needs to exit somewhere.  The lowest resistance path is through the stack, so that's where most of the air will flow.  The big difference is that you are pushing the air into the ER and out though the stack, rather than pulling it up the stack.  Both work, but the unfortunate side effect of pressurizing the ER is that you also push some amount of air into and through the boat.  That both heats the boat, which sometimes can be good, but it has the undesirable effect of also pushing ER odors into the boat as well.  By sucking air up the stack, the ER runs at a slight vacuum and any fumes or smells get pulled out of the boat.

This has been a full two years of head scratching, measuring, fabricating, measuring again, failing, trying again, and finally tripping over the final piece of the puzzle.  The key to this approach is the stack cavity and it's ability to carry a lot of air.  It may not exist on other boats, or be as easily used as on the 55/60 model Nordhavns.  But hopefully this will help others think through how their ventilation air flows, where it might be restricted, and take a more scientific approach to improving it.  After all, science always improves your luck....

Tuesday, August 23, 2016

Gravel Barge Flips in Seattle Ship Canal

Wow, a gravel barge just flipped in the Seattle ship canal.  It was tied up at Salmon Bay Sand and Gravel, actually three barges side by side.  The middle one flipped and come down on top of the outermost barge.  They were loading or unloading at the time, so presumably some excessive imbalance caused it.  The two entangled barges are drifting free and appear to have come up against a pier.  We are directly across the canal and heard it and have been watching since.

The police on scene reported to the CG over VHF that all people were accounted for, no injuries, and no fuel spills.

Monday, August 1, 2016

Auto pilot 2.0 - Furuno NavPilot 700

Almost 18 months ago in this article on my Electronics 2.0 refit, I mentioned that I was moving from a Simrad autopilot system to a Furuno NavPilot 700 series AP.  Here's the brief explanation that I gave:

Dual Furuno NavPilot 700 auto pilots.  Unlike the Simrad pilot, you can actually build up two independent pilots using the Furuno gear so you really have a hot standby.  When I was first evaluating the Simrad gear, I sketched out a dual pilot system and Simrad told me in no uncertain terms, confirmed in writing, that it would work as I expected.  Well, after hours and hours and lots of swapped out equipment, I can tell you with certainty that it DOES NOT work.  Simrad now confirms that it doesn't work, and has never been designed or tested with dual pilots in mind.  Anyway, the Furuno pilots let me finally build the dual system that I wanted from the start.

I've been using it for a while now (probably 3500 miles) and can give a good report on its performance and capabilities, at least in the context of a trawler.

As you can see in the excerpt above, having a redundant auto pilot was an important objective for me.  Actually, our goal has been redundancy for all critical systems, and I certainly consider an AP to be one.  The fatigue level for anyone standing watch is substantially different (like no comparison at all) if you are hand steering vs just monitoring what's going on and watching for logs etc.  From day one the goal was to have two completely independent auto pilots from the control panel all the way through to the steering pump.   That way, if any component fails, you can switch over to the other pilot and keep on going.

I'll save you the drama, but the Simrad pilot can't do that - at least not without a lot of hassle and external switching over from one pilot to the other.  That, plus a hand full of other issues with the
Simrad pilot caused me to switch to the Furuno NavPilot 700.  Doing so was a risk since I had no idea whether I would be making an improvement, or just trading one bag of problem for a different bag.  It turns out to have been a big improvement, but also included a small bag of new problems.

Here's what I like about the NavPilot 700:
  1. All the components that make up the NavPilot are NOT interconnected via NMEA 2000 (N2K).  Connecting everything via N2K like Simrad does it is really nice in theory, but given all the issues with N2K, I now favor segregating critical components whenever practical.  The control panels daisy chain together via CAN bus, but it's a closed and dedicated CAN bus used only by a single NavPilot.  The rudder reference sensors are also directly hardwired to the NavPilot computer.  This means more wires running back and forth through the boat, but it also means a more dependable system.  Pulling wires is something you suffer through once, yet dependability you enjoy every time you run the boat.
  2. Three different control panels are available, along with at least one wired remote.  This gives lots of flexibility to set things up in a way that works best for you, and that fits your available space.  In my case, I was severely constrained for dash panel space, and the small control panels allowed me to fit two with out a problem.
  3. All the control panels have a roto dial control knob.  When the pilot is automatically steering to a particular heading (Auto mode), the dial makes it really easy to adjust the heading in both small and large increments.  I find this much easier than just having buttons to press.
  4. The NavPilot computer supports a variety of interfacing options, not just N2K.  This gives flexibility, plus it provides a way  to have an alternate control path from the chart plotter running Coastal Explorer to the NavPilot.  I now have things set up so the pilot will continue to follow my route even if N2K fails completely.
  5. The NavPilot has a very good (one of the best I've seen) way to manage source selection for all the external instruments used by the pilot.  For each one (GPS, Heading sensor, etc), you can select up to three different devices, and prioritized the order in which the pilot should use them.  It will then pick the highest priority device, and automatically fail over to the next one on the list if the current device fails.  While I was making adjustments and getting everything to work, I encountered a number of these automatic failovers, and other than an alarm from the pilot alerting you that it changed sources, it just kept on steering like nothing happened.
  6. There are no issues with two pilots on the same network, through I do only power one on at a time.  Two on the same network, even powered on at different times, totally messed up the Simrad pilot.
  7. The NavPilot has a much better set of tunable steering parameters.  There are three different sets of parameters that you can set up for three different sea states.  With a click and a dial, you can select which set of parameters you want to use.  This has worked out really well and has allowed me to fine tune the steering parameters.
  8. And last but not least, the pilot steers the boat very well.  Nordhavns can be hard to steer.  They are very heavy, move relatively slow, and have big stabilizers.  Throw in some sloppy seas, and the pilot works hard.  I was always happy with how the Simrad pilot steered the boat, but I can say that the Furuno pilot is at least as good.  Throw in the selectable parameter groups, and I'd give it a nudge above.

Controls for dual pilots (inactive one covered), and programmable groups of steering parameters

Several different control panels, all with roto-dial controls
That's a lot to like, and as a result I've been very happy with the NavPilot.  However, it's not without its shortcomings.

But, as background for discussing one of the shortcomings, it's worth talking about follow-up steering controls.  Smaller boats don't use these, but bigger boats do.  They are lever controls that essentially give you power steering, and very fast power steering at that.  There are two flavors; Follow-up, and Non-Follow-up.

A Follow-up (FU) lever is the easiest to understand.  It's a lever or knob that is used to position the rudder.  Center the lever and the rudder centers.  Turn to 15 deg port and the rudder goes to 15 deg port.  Turn hard to stbd and the rudder goes hard to stbd.  It's a bit like a tiller on a small boat and, to its name, the rudder "follows" the Follow-up lever's position.  These are indispensable for maneuvering larger boats where manually turning a wheel just isn't fast enough.  Our wheel, for example, is 10 turns hard over to hard over, and it's not a one hand operation.

Follow-up (FU) steering control

A non-follow-up (NFU) control is a little different, and has three distinct control position.  Centered, there is no rudder movement.  The rudder stays in whatever position it's in.  Hold the lever to the right and the rudder moves to the right.  As long as the lever is held to the right (it spring-loaded and will pop back to center if you let go) the rudder will keep moving to the right until it hits its stop.  Hold the lever to the left, and the rudder starts moving to the left.  So rather than commanding a rudder position as you do with a FU control, you are commanding rudder movement with a NFU control.

Non-Follow-up (NFU) steering control

With that background, here's the list of shortcoming in the NavPilot 700:
  1. Very limited support for Follow-up and Non-Follow-up control levers.  Furuno only makes one very simple follow-up (FU), and resells a couple of FU and NFU levers from other vendors.  But the bigger problem is that the NavPilot only supports a maximum of two FU controls, and even then it doesn't do well.  On our boat we need a minimum of 3 FU controls; one each in the pilot house, port wing station, and stbd wing station.  And there is an argument for one on the fly bridge and another in the stern station.  And even with two, operation is highly problematic.  Each FU control has an on/of switch.  When you switch one on, the pilot enters FU mode and you can steer from the FU control.  That's fine.  But if you turn the other FU control on at the same time, it operates both of them in parallel.  Electrically the control is very simple and consists of nothing more than a potentiometer which is just a variable resistor controlled by a dial.  When you connect two of them together, you get very unexpected steering results and the rudder veers off in an unexpected way.  The only way the boat steers correctly is if you make sure you only enable one FU control at a time - something that my feeble mind has trouble doing.  I'll be steering from the pilot house, then step out to the wing station to make a final approach and dock.  If you forget to turn off the pilot house control and turn on the wing control, the steering goes haywire.  Not good when docking.  This is something that Simrad does much better than Furuno.  With Simrad, you can have an arbitrary number of FU controls, and each has a "take command" button.  When you take command from one station, it automatically disables any other station.  It works much, much better.  This is an area where Furuno really needs to do some work, and it makes it very hard for me to recommend the NavPilot for boats that need more than one or two FU controls.  However, I have come up with a solution that completely solves the problem, though it requires a lot more work to build than I expect most people will be willing to do.  I'll cover this in another article later on.
  2. Surprise, surprise, the NavPilot has some N2K problems.  For some inexplicable reason, it sends out repeated requests to all devices asking for the Humidity PGN.  These requests go out about every 7 seconds and all devices have to stop and respond in one way or another.  It's mostly benign, but totally brain dead and creates a bunch of unnecessary traffic and potential disruption to other important traffic.  I've reported this to Furuno but they have brushed it off and don't appear to have any plans to fix it.
Fortunately, I've been able to work around both these issues and have a pilot that I'm very happy with.

It's probably worth a few minutes to talk some more about the dual auto pilots since that was such an important consideration.  I mentioned earlier that I only run one at a time.  More specifically, each is on a separate power breaker and I have an interlock that ensures only one is powered on at any time.  I'm actually pretty sure that this is NOT necessary, and that I could leave both powered on at the same time as long as I made sure not to activate them both, i.e. I'd always have to be sure at least one is in Standby mode.  But that's subject to operator error, so I prefer the breaker interlock.  The only down side is that if one pilot fails, I need to power up the second before it can be made active.  But that only takes about 30 seconds, so is really not a problem.

Dual Accusteer continuous duty steering pumps

To ensure everything works, I alternate between the two pilots about once a week.  So I'll run on one for a week, then switch to the other for a week, then switch back.  In the pilot house there are two control panels, one for each pilot.  I leave the cover on the pilot that's not in use so it's easy for the operator to know which one is active.  On the fly bridge there is only one control panel connected to pilot #1.  I figure if we are operating from the fly, we can just use that pilot.