Experiment: Boil v. No Boil (Part 2)

pH as a function of time (hrs). Nutrients added at
roughly 24, and 82hrs.

It's data and conclusion time! (or partial conclusions)

Boil
OG: 1.1080 --> FG: 1.0025
pH 5.60 --> pH 3.56

No Boil
OG: 1.1080 --> FG: 1.0015
pH 5.33 --> pH 3.56

Both fermentations were held within 1C of 20C (both had peaks of 21C @ 92hrs from pitch, right around the pH stabilizing). Nutrient additions and aeration were determined by gravity reading, not time after pitch.

Specific gravity as a function of time. Nutrients
added at roughly 24, and 82hrs; aerated at
roughly 24, 44, 58, 68, and 82hrs.

Unfortunately, the data points were not as close as I would have liked, but life gets in the way of things.

Nutrients were added at roughly 24 and 82 hours after pitch, showing a slight change in slope in the gravity graph, and a 'bump' in the pH graph (the other 'bump' in the pH graph at ~44hrs seems to correlate to aeration).

It is clear that the yeast are active well before a drop in gravity, this is shown by the drop in pH before ~24hrs (when a drop in gravity was first detected). This corresponds to the acclimation period which yeast go through commonly called the 'lag phase'; it would seem that instead of active sugar metabolism, they are more concerned with making an adequate environment for themselves at this point.

As to the differences, it would seem that the boiled must had a higher pH throughout the process until the end of fermentation. This is contrary to common belief that boiling (and skimming) reduces the buffering capacity of mead musts, however, this may be an outlier as the honeys chosen for the must composition do naturally have higher mineral and ash contents than other honeys (avocado and lychee are particularly rich in both, the brazilian pepper had a very high pollen content, and the beach plants have a higher mineral concentration than average (White Jr)).
Also the boiled mead fermented a little faster (about 2 days, though the data is not accurate to the hour, hence it's non-inclusion in the graphs).

Both have been cold crashed for 2 weeks at 5C, and have been treated with sulfites (an addition of 1 campden (Kmeta) was added to each gallon container, this method was chosen for its prevalence throughout the mead and winemaking community). They will age for another 9mos before being racked and bottled. Bottle aging will be in 12oz beer bottles with O2 absorbing caps, and evaluation will be held at several points in their lifetimes (hopefully ~1yr, 2yr, and 3yr, but we'll see).

Experiment: Boil v. No Boil (Part 1)

This is a question that tears at the meadmaking community, it separates, polarizes, indoctrinates, and just generally sparks a nice (or not) debate: do I boil my must? The general reason to boil would be sanitization, but the opponents will cite loss of aromatics as too high a price (especially when honey can be diluted and fermented without fear in most cases); but are there other benefits to be gathered from either method? We shall see.

Materials
Four different honeys where used to make a must considered to be average in flavor and aroma:
1. Winter wildflower from Miami (Lip Smakin' Good Honey): mostly avocado and lychee, maybe some citrus, but not much. 2012
2. Summer wildflower from broward and miami coast (Lip Smakin' Good Honey): late season black mangrove, seagrape, dogwood, and dune grasses. 2013
3. Wildflower blended from most of south florida (Bee Natural): citrus notes and general floral character, not too rich or thin. ????
4. Brazilian Pepper from broward and palm beach (Webb's Honey): brazilian pepper honey from 2013 and 2012.
These were mixed at about 4lbs No. 3, 2lbs No. 1, 2lbs No. 2, 1lb No. 4, to make a must of 1.1060 (volume was irrelevant as you will see).
Instruments used and their tolerances:
1. Hydrometer 1.1300-1.0600 ± 0.0005
2. Hydrometer 1.0700-1.0000 ± 0.0005
3. Hydrometer 1.020-0.980 ± 0.001
4. pH Meter MW102 14.00-1.00 ± 0.02
5. Thermometer 150C-0C ± 0.1C
6. Scale 500.00g-0.00g ± 0.02g
7. Scale 11.000kg-0.000kg ± 0.005kg

Procedure
A must of 1.1060 was created as noted above, 5L was separated and left in a sanitized fermentation bucket. The remaining must (>5L) was placed into a kettle and brought to boil, then held at a boil for exactly 15min, with skimming at boil, 10min and 5min, and stirring at boil, 12min, 9min, 6min and 3min. The ramp time to boil was 15min, SG after boil was 1.1225 (showing that the boil was quite vigorous), and the must was cooled from boil (101C) to 44C in 1.25hrs (using passive cooling means). Once the boiled must reached 20C, it was adjusted with distilled water to SG 1.1060.
Booster blanc at a rate of 0.25g/L, tannin FT blanc soft at a rate of 0.05g/L and potassium carbonate at 0.25g/L were added to both fermenters. Yeast (D21) was rehydrated with go-ferm protect for 15min, then 34.1g of this solution were added to each fermenter, resulting in a pitching rate near 5*10⁶ CFU/mL. These processes ensure that the only difference between the two musts is the reactions of boiling, and skimming, on the one.
Both will be given 1g/L fermaid K at the end of lag (SG ~1.106) and another 1g/L at the 1/3 sugar break (SG 1.072) resulting in a YAN level of 200ppm. Both will be kept in a fermentation chamber with ambient set to 20C ± 1C.

Stats
Five minutes after pitching, SG, pH and temp were observed:
Pitched 8 February, 2014 at 20:00
Boil: SG 1.1080 ± 0.0005, pH 5.33 ± 0.02, 22C ± 0.1C
No Boil: SG 1.1080 ± 0.0005, pH 5.60 ± 0.02, 22C ± 0.1C

The increase in gravity is due to additives and the yeast themselves. Note the lower pH of the boiled must, skimming most likely removes some of the ash that is in the honey resulting in less buffering and less alkali material in the must. Another note was that the boiled must was far clearer than the obviously  opaque no boil must.