The Great Homebrew Coolship Experiment

At Homebrew Con 2015, James Howatt of Black Project Spontaneous & Wild Ales presented a seminar on Wild and Spontaneous FermentationThe seminar highlighted a variety of funky topics and was instrumental in educating homebrewers about proper coolship geometry.

For those unfamiliar with the term, a coolship is a broad, shallow, open vessel used to cool wort. A long time ago, coolships were an efficient, de-facto method of cooling boiling-hot wort in traditional breweries (before plate-, counterflow-, or immersion-chillers were even invented). Today, coolships are found in just a handful of breweries (most often spontaneous beer production breweries), scattered between Belgium and the United States. The reemergence of the sour, wild, and spontaneous beer markets have inspired homebrewers to perform spontaneous fermentation using their own homemade coolships!

The Coolship at Brasserie Cantillon in Brussels, Belgium

However, despite good intentions, many homebrewers are incorrectly using geometrically scaled versions of commercial coolships (i.e baking pan). The problem, as Howatt explained during his Hombrew Con seminar, is most scaled-down coolship surface area to volume ratios (SA:V) are an entire order of magnitude greater than their commercial counterparts, translating to too-rapid of cooling. Milk The Funk (MTF) has since expanded on this topic on their Coolship Wiki Page. They explain that a homebrewer should consider not only top SA:V (top of wort exposed directly to ambient air) but also total SA:V (total surface area of wort on all six sides). For example:

Commercial Coolship

  • Dimensions: 10 ft x 10 ft 1.5 ft
  • Fill Volume: 36 bbl = 1122 gal = 150 ft3
  • Top SA: 10 ft x 10 ft = 100 ft2
  • Top SA:V: 100 ft÷ 150 ft3= 0.67 ft2/ft3
  • Total SA: 2 x (10 ft x 10 ft)  + 2 x (10 ft x 1.5 ft) + 2 x (1.5 ft x 10 ft) = 260 ft2
  • Total SA:V:  260 ft2 ÷ 150 ft3 = 1.73 ft2/ft3

Geometrically Scaled Homebrew Coolship

  • Dimensions: 2.5 ft x 2.5 ft 0.2 ft
  • Filled Volume: 9.35 gal = 1.25 ft3
  • Top SA: 2.5 ft x 2.5 ft = 6.25 ft2
  • Top SA:V: 6.25 ft2 ÷ 1.25 ft3= 5.00 ft2/ft3
  • Total SA: 2 x (2.5 ft x 2.5 ft) + 2 x (2.5 ft x 0.2 ft) + 2 x (0.2 ft x 2.5 ft) = 14.5 ft2
  • Total SA:V: 14.5 ft2 ÷ 1.25 ft3= 11.60 ft2/ft3

This example demonstrates that a scaled-down homebrew coolship is easily six to seven times greater than its commercial counterpart with regard to top and total SA:V! A greater top SA:V corresponds to greater wort-microbe exposure and a greater total SA:V corresponds to quicker cooling rate – both of which impact spontaneous inoculation and fermentation.

Instead of homebrewers using a baking pan, Howatt recommends a boil kettle (aka kettleship) to more accurately replicate the commercial coolship SA:V. While I trust Howatt’s rationale (especially since he won GABF medals using 30 gal kettleships), I wanted to prove it myself!

Thus, on January 15, 2017 the Great Homebrew Coolship Experiment kicked off!

Purpose

The purpose of the experiment was to evaluate the cooling duration of two different homebrew coolships; Bayou Classic 1082 82 qt Kettleship vs. Igloo 10 gal Coolership. Twenty gallons of untreated water were brought to a boil and transferred into each coolship. The coolships were transported outdoors on a brisk 34°F afternoon. Kettleship, coolership, and outside temperature was recorded every two minutes (using a BCS-462) until both coolships reached a terminal target temperature of 64°F.

Methods

Since the coolership is plastic, a phone call to Igloo’s customer service department revealed that it is made of polypropylene (PP). PP is widely used for food containers due to its high melting temperature (266°F), lending it suitable for freezer, microwave, and dishwasher temperatures. Despite its wide use, PP (and many other plastics) are being reviewed for their tendency to leach byproducts into food and beverages.

In preparation for the experiment, kettleship and coolership SA:V calculations were computed and verified using the MTF Coolship SA:V Calculator.

Bayou Classic 1082 82 qt Kettleship

Bayou Classic 1082 82 qt Kettleship

  • Dimensions: 1.58 ft (diameter) x 1.42 ft (height)
  • Filled Volume: 10 gal = 1.34 ft3
  • Top SA: π x (1.58 ft ÷ 2)2 = 1.96 ft2
  • Top SA:V: 1.96 ft2÷ 1.34 ft3 = 1.46 ft2/ft3
  • Total SA: 2 x π x (1.58 ft ÷ 2) x 1.42 ft + π x (1.58 ft ÷ 2)2 = 9.00 ft2
  • Total SA:V: 9.00 ft÷ 1.34 ft3 = 6.72 ft2/ft3

 

Igloo 10 gal Coolership

Igloo 10 gal Coolership

  • Dimensions: 1.08 ft (diameter) x 1.63 ft (height)
  • Filled Volume: 10 gal = 1.34 ft3
  • Top SA: π x (1.08 ft ÷ 2)2 = 0.92 ft2
  • Top SA:V: 0.92 ft÷ 1.34 ft= 0.69 ft2/ft3
  • Total SA: 2 x π x (1.08 ft ÷ 2) x 1.63 ft + π x (1.08 ft ÷ 2)2 = 6.46 ft2
  • Total SA:V: 6.46 ft2 ÷ 1.34 ft3 = 4.82 ft2/ft3

Comparisons were also made (after-the-fact) to the MTF Commercial Coolship Table (which has been consolidated and converted to imperial units, below):

[table “” not found /]

The kettleship and coolership had similar top SA:V to Jester King and Lindemans, respectively, indicating that wort-microbe exposure was appropriate. Their total SA:V were greater than any brewery, indicating that cooling rate may be too rapid. According to the MTF Coolship SA:V Calculator, the estimated cooling time for the kettleship was 8.83 hours, which seemed appropriate. I assumed the coolership would take longer to cool due to the cooler insulation, somewhere around 12-14 hours.

Ten gallons of untreated Holland, MA well water was heated to boiling and transferred into the coolership. The coolership was placed outside. The process was repeated for the kettleship. Once outside, coolership, kettleship, and outside temperatures were recorded every two minutes using a BCS-462 until terminal target temperature was reached (64°F). Final volume was recorded.

Results

BCS data was imported to Microsoft Excel and a time-temperature scatter plot was generated for the kettleship and coolership:

 

Time-Temperature Scatter Plot

Critical calculations were also generate using the BCS data, including cooling duration, change in temperature, evaporation, average outside temperature:

Critical Calculations Chart

Discussion

As anticipated, the kettleship cooled faster than the coolership. This can be attributed to the coolership insulation and the larger kettleship diameter (which resulted in 2X greater evaporation rate).

The kettleship required 7 hours and 10 minutes to cool 10 gallons of water from 205°F to 64°F. The wide kettleship diameter resulted in 1.25 gallons (13%) of evaporation, which corresponds to an increase of 7 gravity points for a 1.050 SG wort (0.8% ABV).

The coolership required 17 hours and 34 minutes to cool 10 gallons of water from 196°F to 64°F. The lower starting temperature can be attributed to the initial heat absorption of the cooler insulation. A higher starting temperature would have extended the cooling duration even longer. The narrow coolership diameter resulted in 0.7 gallons (7%) of evaporation, which corresponds to an increase of 4 gravity points for a 1.050 SG wort (0.4% ABV). The coolership also experienced a lower average outside temperature due to the longer cooling duration that extended into and throughout the night – a higher average temperature would also have extended the cooling duration longer.

The MTF Coolship SA:V Calculator was not very accurate in estimating the cooling duration for the kettleship and even worse for the coolership. This may be attributed to the differences between wort and water with respect to density and specific heat. The calculator’s heat transfer coefficients may be an average of metals and coolship thicknesses, not taking into consideration plastic coolers. It also does not take into consideration ambient humidity, which could presumably impact evaporation rate and therefore cooling rate. All in all, the calculator was in the ballpark of the observed cooling duration and may be a useful tool in assessing qualitative (not quantative) performance of a kettleship.

It is unclear as to whether the longer cooling duration of the coolership is beneficial. It may give wild bacteria a longer head start to replicate in the cooling wort, out-competing the wild yeast and contributing to a more acidic (or funky?) beer. In contrast, the longer cooling duration may allow for the wort to collect a greater amount of wild yeast, allowing for a shorter lag time to fermentation, creating to a less acidic/funky beer. Anecdotally, Hollambic #1 was cooled in a coolership and was noticeably more acidic compared to Hollambic #3 (which was cooled in a Kettleship) – however hopping rates varied significantly between the two batches which may have contributed to the perceived acidity. Based on the suggested target cooling rate (8-12 hours) from Belgian lambic production methods, I would recommend the used of the kettleship over the coolership for spontaneous beers, though both can be used to successfully inoculate spontaneous beers. The kettleship also avoids any debate regarding plastic leaching byproducts!

Limitations of this experiment include the use of water instead of wort, which was selected out of convenience. Since wort has a greater specific heat than water, it is feasible that if this experiment were repeated with wort, the cooling duration would be shorter than the reported values. The experiment was also a n=1, therefore it should be replicated before definitive conclusions can be drawn from the data.

BCS-462 Dashboard for Temperature Monitoring

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.