Problem with pre-boil and post-boil gravity estimates

Beersmith's reported pre-boil and post-boil gravity estimates appear to be off. For a simple test, I have created an all-grain recipe in beersmith 2 using an unmodified equipment profile:

Type: All Grain
Equipment: Pot and Cooler (10 Gal/37.8 L) - All Grain
Batch Size: 5.50 Gal
Ingredient: 10 lbs Pale Malt (2 Row) US

The following fields are calculated by beersmith:
Est Pre-Boil Vol: 7.89 Gal
Est Pre-Boil Gravity: 1.039 (39 gravity units)
Post Boil Vol: 6.76 (found in the Equipment profile page)
Est Original Gravity: 1.048 (48 gravity units)

I would expect the following formula to hold true:
(Pre-boil Vol) x (Pre-boil gravity units) = (Post-boil Vol) x (Post-boil gravity units)

Plugging in beersmiths values, I get:
7.89 x 39 = 6.76 x 48
307.71 = 324.48

In order for this to work, the Est Original Gravity needs to be 307.71 / 6.76 = 46 (or 1.046 SG).

My calculation can be verified using this calculator (the second one on the page):
http://www.brewersfriend.com/dilution-and-boiloff-gravity-calculator/

Why is beersmith reporting an est original gravity of 1.048 instead of my calculated value of 1.046? Am I missing something here?

You need to adjust the post boil volume for water expansion at boiling (4%) and then the numbers come right into line.

OK, I see. Thanks for your answer. That means that the pre-boil volume doesn't have any temperature expansion applied to it - it is the volume at room temp. The tooltip on the "Boil Volume" field misled me. It reads: "The volume of the wort at the start of the boil".

Yeah, it is a bit of a head scratcher.  I played around with the expansion rate and the numbers followed it cleanly.  Hand calculating the extraction, I get the numbers pre-boil based upon cold volume.  Theoretically, you never actually measure the post boil volume as cold, because after you chill it the program is factoring in chill and trub losses.  But then again, I never measure the volume pre-boil as cold either.  In adjusting my equipment profiles for my actual results, it seems to take this into account as I usually hit both pre-boil and post boil volumes and gravities without issue.

I'm still trying to wrap my head around what beersmith is doing. I tried the following:

Set boil off to 0 gal
With tot efficiency set to 75%, I get the following:
Estimated pre-boil gravity: 1.048
Estimated OG (post-boil): 1.050

This is a 4% difference, but temperature expansion should have no affect here since all gravity measurements should be calibrated to room temp. We start with wort at 1.048, do nothing, and end up with 1.050? I wish brewing was that easy!

Unless I'm doing something really dumb, this looks like a bug to me.

I can work around this problem (which I'm now convinced is a bug) by disabling the "Cooling Shrinkage %" (setting it to 0), and adding the 4% shrinkage amount to the "Loss to Trub and Chiller" field.

Now both the "boil volume" and "post boil vol" are volumes as measured at boiling temperature, and the pre-boil and post-boil gravities come out as expected.

Using this work-around in the test recipe of my original post:

Pre-boil volume: 7.89
Est Pre-boil SG: 1.041
Total GU pre-boil = 41 x 7.89 = 323.49

Post-boil volume: 6.76
Est Post-boil SG: 1.048
Total GU post-boil = 48 x 6.76 = 324.48

I don't focus on the post-boil volume as much as the post boil gravity.  I spent the first few brews doing careful measurements of the variables during the process: initial water volume, grain weight, pre-boil volume and gravity, post boil gravity, fermentor volume and loss to trub/chiller/remaining volume.  I ended up changing the water absorption by grain and boil off rate to match actual results.  The actual volume post-boil becomes lost in the process, but could be back calculated using the ending volumes for the fermentor and losses.  Since I have made these adjustments, the program works remarkably well and very consistently predicts my actual results both for pre- and post- boil gravity and wort volume from mashing/sparging, fermentor volume and trub losses.  Any calculations based upon hot versus cold volume get absorbed in the tuning of the profile parameters.

Yes, I agree that you can play with various equipment properties to get estimated gravity readings to work consistently with actual brew-day measurements. However, my point is that doing that is still a workaround to what appears to be a flaw in the calculation beersmith uses to determine estimated pre-boil gravity when shrinkage % is activated.

Moving the % shrinkage into the trub loss field makes my actual brew-day measurements line up nicely with beersmith's estimates. Since you already have your system dialed in, you don't have to worry about it. However, on my brew day last week, I hit the estimated pre-boil gravity dead on, but missed the post-boil SG since the estimated pre-boil gravity beersmith calculated was too low. If I knew that the pre-boil gravity needed to be higher, I could have added some DME to correct for it.

Hopefully this workaround will fix my problem, but I'll need to do a couple more batches to see how the numbers line up.

As a former process control engineer, I would say it differently.  The software is a modeling system for the brewing process.  The actual results for key points in the process are fed back into the software in terms of settings which allow the software to better predict the intended outcome for future brews.  What matters is really the input variables (grain, hops, water, additives, yeast, boil time, temperatures) and the process specific constants (water loss to grain, extraction efficiency, evaporation rate, specific heat of equipment, hop utilization) produce a consistent result: beer (OG, color, est IBU, FG).  If you do not modify the process specific constants to match your system, the software may be an OK predictor of your results, but never something you can rely upon entirely.  Internally derived values, such as post boil volume, are benchmarks for process control or information.  You will note in the software that it does not have a user input for actual post boil volume since that measurement is not critical to the output measurements.  It is a benchmark measurement given at prevailing conditions (100C).  The actual result that is critical is the amount of chilled wort (post boil volume less shrinkage) at the measured specific gravity (at 20C).  For this the software is consistent.

I see what you're saying. As long as my mash efficiency never varies by too much, and I have properly tuned my equipment profiles, I will always get the correct FG at the end of the day. I think my issue is that I actually want to know some of the benchmark values since my mash efficiency might change unpredictably from brew to brew.

For example, let's say my usual mash efficiency is 85% and I formulate a recipe with that efficiency to arrive at my desired FG. However, on the brew day I make a mistake such that my efficiency is not what I expected - maybe I only get 75% mash efficiency. When I check my pre-boil gravity, I notice that it is too low compared to my recipe's estimated pre-boil gravity. In this case I can correct the mistake by either adding some DME or boiling longer than planned. Since I checked the pre-boil gravity, I should never be surprised by a lower than expect FG at the end of the boil.

In order to do this I need to know that beersmith's estimated pre-boil gravity (and volume) benchmarks are calculated properly and displayed in a useable format. Alternatively, I can manually calculate my predicted pre-boil gravity, but I would prefer to leave all the calculations to beersmith.