Monday, March 30, 2015

How much is it?

I just did a post which talked about establishing what the right amount of glycol in a holding plate solution was.  Well, OK, now that we know what we want to have in there, how do we determine what we actually have?  I suppose one possibility would be to simply empty the holding plate and refill it with a solution of known concentration, one that we just made up by careful measuring.  Yeah, that would work.

But what if we just want to add a little water or glycol to what we already have?  For that we would need to be able to measure the concentration in the solution. 

Hand-held refractometer
This is the tool for that:  It is a refractometer - it measures the refractive index of a liquid.

What is that?

All transparent substances slow the passage of light thru them somewhat - some more than others.  The refractive index is the ratio of the speed of light in a vacuum compared to the speed of light in the transparent substance.  So, if I tell you that the refractive index of a particular glass is 1.33, that means that, yes, light travels thru that glass only 3/4 as fast as in a vacuum.  You have witnessed refractive index differences when you, for example, mixed water and vodka, or dove in a place where fresh water and salt water are mixing (at Shilshole, for example).

But for our purposes, it is enough to know that the refractive index of a water/propylene glycol mixture changes in a predictable way with the concentration of glycol.  We don't even have to know the details of that change because the manufacturer has taken that into account in the preparation of the scale inside the instrument. 

All that remains is for us to obtain a drop or two of the solution and put it onto the prism covered by the clear plastic flap, and look thru the lens at the other end of the instrument, for a view like this:

Approximately 33% propylene glycol shown
But there is a catch (isn't there always?).  Before I sample the holding tank solution, I have to completely defrost the freezer.  If I don't, some unknown quantity of the water in the system will be frozen out, which would skew the results in the direction of increasing concentration.  So the refractometer stands ready for duty, waiting for a freezer defrost event.

Who knows when that will be?


Monday, March 23, 2015

Destination: Center Island, San Juan Islands

In Lopez Sound, tucked into the bight formed by the spit on the south end of Decatur Island, lies Center Island.  

Lopez Sound
Excerpt from Chart 18429
The island is all privately owned and is large enough to have an airstrip.  Tho there is no public access to shore, Center Island has one very important attribute for cruisers in the San Juan Islands:  it is the eye of the storm.  On multiple occasions when the wind was howling out of the west, down the Strait of Juan de Fuca and Hunter Bay, we found almost calm conditions anchored off the north east shore of Center Island.  And then there was the time when we came through Lopez Pass with 25+ kt on our stern and anchored in the same spot, again in calm conditions.

Center Island
Depths in fathoms
Approaching from Lopez Pass, we clear the last of the chain of small islands north of the pass, turn north, and round Center Island on its western side.  The area between Center Island and the Decatur spit is quite shallow south of the cable area, and so we avoid it in all but the highest of tides.

Approaching from the north there are no surprises.  We anchor in 4 fathoms, close to the island and just north of the edge of the cable area. 

Calm Morning
That is Center Island to the right of center, viewed from the north

It's not always this calm at Center Island, but it is a beautiful anchorage.


Monday, March 16, 2015

How Much Is Enough?

The other day I got to wondering if I could improve Eolian's freezer - to make it colder, or to make the cold last longer, by fine-tuning the amount of propylene glycol in the holding plates.

But first, a word of explanation.  Eolian has two separate compartments for the refrigerator (one holding plate) and the freezer (two holding plates), but both are cooled by a single compressor.  The refrigerator has a thermostat in it, which closes a valve in the refrigerant line when the setpoint is reached.  The freezer on the other hand has no valve, and continues to chill - until the compressor shuts itself off.

So I decided to do put my education to work and do a little Chemical Engineering.

You may remember that in a previous post I discussed the phase diagram for the propylene glycol-water system earlier.  I suggest you go back and read that one now, because I will proceed presuming you have...


But this diagram unfortunately does not have any actual temperatures or concentrations delineated on it.  So I prepared an expanded version of the portion to the left of the eutectic point from some published data:

Portion of Propylene Glycol-Water Phase Diagram

But even this covers too much territory.  What we are interested in is just this portion:

The Part We Care About

As I mentioned in that previous post, I spoke at length with a Technautics sales rep at a Seattle boat show, and he recommended a 15% solution for freezer holding plates.  I don't know what the concentration in ours is (that is something I should rectify...), but let's assume that 15%.  Here it is plotted on the diagram:


The vertical pink line describes a 15% concentration.  Let's follow what happens as the temperature is dropped, moving down the pink line at the 15% concentration.  At first, the liquid just gets colder.  But at a temperature of 22°F we meet the line.  At this point, the first freezing occurs.  And what freezes out is ice - pure water ice.  Because of this, the amount of water remaining in the liquid phase is reduced, increasing the concentration of propylene glycol. 

Now let's look at the situation at 15% glycol cooled to 10°F... the intersection of the two pink lines.  By this point, we have frozen out a substantial amount of ice, removing water from the liquid phase.  In fact, the concentration of glycol in the liquid phase has followed the black curve down, and is now about 27% (the concentration of the total mixture has not changed - it is still 15%).  And the Lever Rule tells us that the total mixture is a slush of 12/27 = 44% ice and 15/27 = 56% glycol solution.

Now why did I draw the horizontal pink line at 10°F?  Because that is the lowest temperature our compressor can reach with our current refrigerant.    A low pressure switch cuts it off.  The only way to decrease this endpoint would be to change out the pressure switch (not recommended - the manufacturer decided that this is the lowest pressure at which the compressor should operate), or to change to a refrigerant that has a higher vapor pressure.  In fact, we have already done that; we replaced the original R-12 with R-409a, which indeed has a higher vapor pressure.  So 10°F is as cold as it can get in our freezer.

So that's one point settled:  The low temperature is 10°F.  Now how to maximize the time that the holding plates can maintain the cold?

 It is important to keep in mind that the holding plate is going to operate over a range of temperatures.  In our current case, once the compressor shuts off, the first bit of ice begins to melt, at a temperature of 10°F, diluting the glycol solution.  As more ice melts, the temperature rises, until finally when the last bit melts, the temperature of the mixture has reached 22°F.

To then summarize, with a 15% solution, the holding plates will operate over a temperature range of 10°F -> 22°F, and have a cooling capacity representing 44% of the total original solution.

We could increase the amount of glycol in the plates... say to 20%.  The plates would then operate over a temperature range of 10°F -> 18°F, and have a capacity representing only 26% of the total original solution (calculation left as an exercise for the reader).  So we would give up nearly half of our capacity to maintain temperature in exchange for a 4°F drop.  Doesn't seem like a fair trade.

OK, how about if we go the other way?  What happens if we decrease the glycol concentration?  Let's try 10%, the concentration that the Technautics rep recommended for refrigerator holding plates.  Now the holding plates operate over a range of 10°F -> 26°F, with a capacity representing 63% of the original solution.  Wow!  We increased our capacity by 50%!  But we have to tolerate an upper temperature of 26°F.  I think that is uncomfortably warm for a "freezer"...  many foods will not be solid at this temperature because their water content is saturated with eg. sugar - like ice cream.

Conclusions

  • Technautics has it right.  The 15% concentration provides a good balance between capacity and temperature range.
  • And I need to assess the glycol concentration in our freezer holding plates.

* Someone is almost certain to suggest that the glycol concentration be raised to 30%, or even more, to get a lower freezing point.  That would certainly accomplish that, but because of the limitations of the compressor and refrigerant, we would never reach that freezing point.  In fact, the holding plates would have zero capacity the way it was assessed above - they would contain nothing but liquid at the compressor cutoff, and would begin warming immediately.  The magic of the holding plate is the ice - melting a gram of ice at 32°F to a gram of water at 32°F absorbs 80 times more heat than raising the temperature of that same gram of water from 32°F to 33°F. 

Monday, March 9, 2015

Chemistry Experiment: Results

When I did this experiment, I promised you the results. Here they are:

Before

After

The benzalkonium chlorides definitely killed the lichen.  But it did take a mild scrubbing with a brush to remove the corpses.  Please note that previous scrubbing with the same brush and bleach had no effect on the spots.  I apologize that I used a washer for the sizing comparison - I didn't have a dime in my pocket.  The washer is a little bigger than a dime and a little smaller than a nickle.

I also applied the 5% benzalkonium chloride solution to a portion of our dodger canvas.  Please note that this canvas is more than 10 years old, and has lost most of its water repellant qualities.

Untreated

Treated
The "After" photograph does not do the results justice.  The mold, mildew and algae are all dead.  A couple of rainstorms washed the bulk of the corpses away, except at the seams.  But more rain is coming (of course).

Conclusion

This stuff works!  And Jane tried it on the moss on our driveway at a 3% concentration, as another experiment.  It worked there too - the moss is all dead.  In fact, using a 5% solution was probably overkill - my next application on Eolian's canvas will be at 3%, but I think I'll retain the 5% level for the tough-to-kill lichen on the decks.

Monday, March 2, 2015

New Spreaders For The Mizzen

Rot.
It's funny, isn't it - how something like this can prey on your mind - kind of subconsciously, but still gnawing away just below the surface.  Just like the rot - slow  but relentless.

Yes, I've known about the rot in our mizzen spreaders for a while now.  And I have been intending to deal with it for all that time, but the "right" time hadn't occurred - until now.

So I tied off the mizzen halyard to act as an auxiliary cap shroud and climbed the mast with my tools.  Removal of the spreader was far easier than my worries had been leading me to anticipate, even tho it had been up there for 37 years, untouched and uncomplaining.

More rot at the root.
I had decided that the replacement spar would be made from  pressure treated lumber (like the second-generation bowsprit) so that rot would never again be an issue.  Now if you have ever hoisted up pressure treated lumber at the lumber yard, you know that it is heavy.  And heavy is not good aloft on a sailboat.  But most of that weight is water:
To pressure treat lumber,
  • it is put into a sealed vessel,
  • the vessel is evacuated, removing nearly all the contained water in the wood as well as air trapped in its porosity.
  • Then after a suitable degassing period, the water-based treatment solution is admitted to the vessel, and it is pressurized, forcing the solution deep (well kind of - see below) into the wood.
That's why the wood is so saturated with water when you buy it.   But if you simply let it stand in a reasonably dry place for long enough, it will return to a more "normal" moisture content - and much lighter weight.  In fact, the replacement spreader is actually lighter than the old one (which of course does have a lot of water in it due to the rot).

Well it only took a couple of hours with a skill saw, power plane, belt sander, and a router to duplicate the shape of the old spreader:

 

If you've ever cut into a pressure treated board, you know that despite the violence of the treatment process, the treatment does not reach the core of the board.  And since the spreader tapers from 1-1/2" thick at the root to 1" at the tip, and from 5" to 1" side to side, a lot of wood was removed from the surface and from one side.  Tapering the thickness and cutting away the sides revealed wood that was only lightly treated.  What to do?

People who handle pressure treated lumber day in and day out (building decks, for instance) have always had to deal with the cut ends of boards, where the untreated core gets exposed.  Long ago they found the answer:

Want to rot-proof some wood?  This is just the ticket
 
This stuff, which you can find now that you know its name, is just the ticket.  You just paint it on and wait for the solvent to evaporate.  Wear rubber gloves - its not nice stuff.

Add a coat of epoxy to permanently seal the surface.
Bed the hardware in polysulphide.

A couple coats of paint, and et voilĂ !
And now, even tho I still have one more spreader to go, it feels like an invisible weight has been lifted from me. The subconscious mental anxiety that was silently eating away at my contentment, my satisfaction, my peace, is now gone, just like the rot.