Honneyman method for ABV
Claude Jolicoeur
ABV is getting quite crowded...
I have looked a little more deeply into the Honeymann table. I was
wondering in fact if I could rebuild the table from standard
alcoholometric tables such as the OIML table that we can find here:
http://www.oiml.org/publications/R/R022-e75.pdf
And the answer is yes - i.e. the Honneyman table is a simple reworking
of an alcoholometric table that would give the ABV of a pure water-
alcohol mixture in function of the SG (60F/60F).
This does makes sense... If you have a 5% ABV mixture of pure water
and alcohol, its SG(60F/60F) will be 0.9929 (note this is from
interpolation in the OIML table). Now, remove the alcohol and replace
it with distilled water, the SG(60F/60F) will be 1.000, for a
difference in SG of 0.0071. Repeat this for different values of ABV
and present it the other way around to get ABV in function of the
variation of SG and you obtain the Honneyman table, exactly.
The assumption in the Honneyman method is then that this difference in
SG (i.e 0.0071) will be maintained for any cider/wine/beer that has
5%ABV. In other words the soluble solids that raise the SG of a 5% ABV
cider to say 0.999 or 1.002 will raise it by exactly the same amount
once the alcohol will have been removed.
This is of a disarming simplicity!
Claude
Duncan
Duncan
Claude Jolicoeur
> Honneyman's assumption is that when a given mass of solute (the cider solids) dissolves in water or a water ethanol mixture the resulting volume will be the same. Almost certainly it will not however but I have no idea as to the magnitude of the discrepancy, perhaps it is trivially small.
exactly 1 litre - one is filled with pure water and the other filled
with a mixture of alcohol and water. Now add to the 2 bottles the
soluble solids that make a cider different from a pure water-alcohol
mixture, i.e. some sugar, acid, tannin, flavours etc. If we add the
same quantity of soluble solids in each bottle, I don't see why the
density wouldn't change by the same amount in both. Well, there could
be some slight second order effects - I guess some tests are required!
>Sorry if this is clutching at straws but I like the idea of Lloyd being right!
been aware of his work. Then if he has decided to publish a newer
method, it might have been because there are some flaws in the Lloyd
approach. Note also that Lloyd gives absolutely no scientific
justification for his procedure... We need Honneyman's paper to see
his justification for his method!
I did a few side-by-side calculations to give us an idea of the
difference between the two.
Let us assume a finished cider SG 1.000, and with alcohol replaced by
water, SG 1.007
Honneyman: difference is 0.007, i.e. alcohol same as a water-alcohol
mixture with SG 0.993, which gives 4.90%ABV.
Lloyd: ratio is 1.000/1.007 = 0.99305, which gives 4.86%ABV
Difference between the 2 is 0.04%ABV, which is rather negligible.
Now for a cider which has some residual sugar, let's assume finished
cider SG 1.020, and with alcohol replaced with water 1.027. Honneyman
gives same result as above,
Lloyd: ratio is 1.020/1.027 = 0.99318, which gives 4.76%ABV
Difference has increased to 0.14%ABV, still quite small.
In conclusion, for all practical purposes in cidermaking, the 2
approaches are equivalent. For ice cider however, the difference could
become more noticeable. A typical ice cider has a finished SG of the
order of 1.060. Assuming SG raises to 1.074 when alcohol is replaced
with water, a difference of 0.014.
Honneyman: alcohol would be the same as a water-alcohol mixture with
SG 0.986, which gives 10.55%ABV.
Lloyd: ratio is 1.060/1.074 = .98696, which gives 9.73%ABV
The difference between the 2 is now close to 1%ABV, this is becoming
quite significant.
Claude
MikeB...@aol.com
> Honneyman's assumption is that when a given mass of solute (the cider solids) dissolves in water or a water ethanol mixture the resulting volume will be the same. Almost certainly it will not however but I have no idea as to the magnitude of the discrepancy, perhaps it is trivially small.
Duncan Galletly
once the alcohol will have been removed".
Claude Jolicoeur
> The only thing that can interefere with the calculation is if some other
> substances evaporate along with the ethanol. Or am I missing something?
measuring the alcohol would not be accurate. And since that
distillation method is considered as the most accurate, we may I guess
consider that only ethanol and water do evaporate.
Further, the classical method to measure the soluble solids is to
evaporate all the liquids and then measure the weight of the dried
solids. If these did evaporate in part, we would not have an exact
measure this way.
Claude
Nick Bradstock
it is the legally defined method. Whatever drawbacks or faults it may have,
the result given by distillation done in the specified manner by an
accredited analyst is legally binding for duty purposes (subject to the
specified condition - see next) - and label declarations too I believe.
Some adjustment can be made for interference by substances that render the
method inaccurate - but this would have to be proven to HMRC's satisfaction.
Distillation would be used to decide a dispute.
Any other (respectable!) method of analysis is acceptable as long as it can
be shown to refer on a repeatable basis to results found by distillation.
Best
Nick
Sn 26 (of HMRC Notice 162 - Cider production, Sept 2010):
Distillation analysis: method of determining the strength of cider.
This method is referred to in section 9.
If there is a dispute over strength our officer may take samples of cider
which will be analysed using the following method:
• take a representative sample and, after clearing it of sediment and gas in
an approved manner, a definite quantity by measure at the temperature of
20°C is made
alkaline by the addition of calcium hydroxide, and then distilled
• make up the distillate at the temperature of 20°C with distilled water to
the original measure of the quantity before distillation
• ascertain the strength of the distillate by determining its density in air
at the temperature of 20°C by means of an approved pycnometer used in an
approved manner, and
• the strength of product is the percentage of alcohol by volume in the
'Laboratory Alcohol Table'* which corresponds to the density determined by
this method. Where the density falls between two consecutive numbers in the
table the strength will be determined by linear interpolation.
Where the result ascertained by this method is rendered inaccurate by the
presence of substances other than alcohol that method shall be adjusted in
such a manner as may be approved by HMRC for the purpose of producing an
accurate result.
*'Laboratory Alcohol Table' means a table of which a copy, signed by the
Chairman of the Commissioners and identifying it as relating to the Spirits
Regulations 1991, has been deposited in the office of the Queen's
Remembrancer at the Royal Courts of Justice.
--
You received this message because you are subscribed to the Google Groups
"Cider Workshop" group.
To post to this group, send email to cider-w...@googlegroups.com.
To unsubscribe from this group, send email to
cider-worksho...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/cider-workshop?hl=en.
Andrew Lea
> MikeBisp wrote:
>> The only thing that can interefere with the calculation is if some other
>> substances evaporate along with the ethanol. Or am I missing something?
>
> Mike, if this would be the case, then the distillation method of
> measuring the alcohol would not be accurate. And since that
> distillation method is considered as the most accurate, we may I guess
> consider that only ethanol and water do evaporate.
Not quite true, I'm afraid. Of course many of the flavour volatiles
distill over but they are reckoned in parts per million hence too little
to affect the results. More significantly, acetic acid distills over and
I have been told that lactic acid can partially steam distill as well.
For this reason the determination of alcohol by distillation is usually
done on a solution previously made alkaline to neutralise the acids (see
Nick's note). But such an alkalised residue would be no good for
determining 'extract' nor for the Honneyman method.
>
> Further, the classical method to measure the soluble solids is to
> evaporate all the liquids and then measure the weight of the dried
> solids. If these did evaporate in part, we would not have an exact
> measure this way.
Again, not quite true. To get an accurate measure of 'extract'
gravimetrically you have to do the evaporation very gently on a
waterbath. Direct heating can lead to carbonisation which may lead to
loss of mass and may also possibly lead to change of SG by formation of
new Maillard compounds from sugar degradation at high concentration.
Duncan has already mentioned the volume contraction that takes place
when alcohol and water are mixed. All these effects may be small
(except the effect of volatile acid, depending on the sample) but they
need to be recognised as potential sources of error.
Andrew
--
Wittenham Hill Cider Page
http://www.cider.org.uk
Claude Jolicoeur
> More significantly, acetic acid distills over and
> I have been told that lactic acid can partially steam distill as well.
> For this reason the determination of alcohol by distillation is usually
> done on a solution previously made alkaline to neutralise the acids (see
> Nick's note).
affect distillation results and that acidity should be neutralised. I
suppose this would be a more important factor in spanish ciders which
contain more acetic/volatil acidity.
Question - how much volatile acidity may a cider contain before being
considered awful, and how could this amount of acidity affect the
result of distillation (assuming distillation without neutralisation
of acidity) and of Honneyman?
> Again, not quite true. To get an accurate measure of 'extract'
> gravimetrically you have to do the evaporation very gently on a
> waterbath. Direct heating can lead to carbonisation which may lead to
> loss of mass and may also possibly lead to change of SG by formation of
> new Maillard compounds from sugar degradation at high concentration.
solids are kept. This would also apply to Honneyman - the slowlier the
alcohol in the cider is evaporated, the more accurate the method would
be...
> Duncan has already mentioned the volume contraction that takes place
> when alcohol and water are mixed.
since it is derived from alcoholometric tables, and these tables exist
because there is volumic contraction - if there was no such volumic
contraction, we wouldn't need the alcoholometric tables...
Claude
Duncan
> Duncan has already mentioned the volume contraction that takes place
> when alcohol and water are mixed.
>Yes, but this is already taken into account in the Honneyman table,
>since it is derived from alcoholometric tables, and these tables exist
>because there is volumic contraction - if there was no such volumic
>contraction, we wouldn't need the alcoholometric tables...
In relation to the effect of solute, can I give a worked example.
Imagine we have the components of a cider, simplistically some soluble solids, and a water – ethanol mixture. So as not to complicate things, lets assume that water and water-ethanol both have a density of 1.0g/ml. They don’t of course but it isn’t relevant to this part of the discussion.
If we take 20grams of our solids and place them into a volume of 990ml ethanol-water we may find (fortuitously) that the resulting volume is 1000ml. The resulting density is therefore 20+990 / 1000 or 1.010g/ml i.e. a SG of 1.010.
Mr Honeyman now comes along and evaporates (gently), all of the solvent away, leaving us back with the original soluble solids, he then begins to add distilled water.
Now if 20grams of the solids expands the
water slightly more than the original alcohol-water solvent, such that 20grams
of solvent and 990ml of water makes up a volume of 1005ml, Honneyman will find
that he only needs 985ml to reconstitute the original 1000ml volume. This time
the density is therefore 20+985 / 1000 or 1.005 i.e. an apparent drop in SG
of 0.005. This error will add to the densitometric variation of the alcohol / water mixture of course, but the latter can be removed from tables, but the 0.005 error remains.
Because this error increases as we increase the mass of solid, and also perhaps as alcohol increases, we possibly need to take starting SG into the calculations, hence proportional change in SG may be better than subtraction.
It transpires that in fact all of this was known over a hundred years ago, that Honneyman probably had little or no role in the development of the method and that modern formulae may provide more accurate analysis.
The original method was described by a French chemist by the name of Tabarie in 1830. He developed tables for calculating alcohol from a comparison of starting SG and SG after evaporation and reconstitution. The original formula used the ratio of the two values but in the 1890’s a Thomas Blunt suggested that the relationship was better described by subtraction. There was some disagreement at the time with subsequent papers by Harvey, Allen and Leonard, providing no clear consensus. Textbooks at the time said that the subtraction method should not be used but it retained its advocates. Lloyd in his cider analyses clearly disagreed however and agreed with the textbooks.
Honneyman was clearly an advocate for subtraction 60 years later, and it seems today the subtraction method is the one most commonly encountered.
Some modern authors have questioned this subtraction method. Hackbarth in the Journal of American Brewing (2009) realised that "This approach assumes that the binary solutions are independent and does not consider ethanol-sucrose interactions or the disruptive effects a second solute has on intermolecular forces." In 2011 Hackbarth in the same journal goes on to provide a more accurate formula.
Hackbarth's criticism of the subtraction method I guess is similar to my problem with it and interestingly the same views were expressed a century before when, in 1897, a theoretical analysis (The Analyst September 1897) noted that “As we have pointed out, the subtraction formula does indeed imply that the substances exert their specific influence on the density of the solution, independently of each others presence; but this supposition is not in accordance with our experiments. The latter show that when solutions of sugar and alcohol are mixed, a slight expansion takes place, the actual density of the mixture being lower than that calculated by the subtraction formula.”
Over 100 years later Tabarie’s method is
still in use but in recent years the formula seems to be undergoing attempted revision e.g. Hackbarth.
Thus Honneyman’s method is actually the method proposed by Blunt in the 1890’s and its problems were recognised at the time. Blunts method (nee Honneyman's) incorporates many errors a) that due to the expansion/contraction of solute, b) the complex effect of distilling off non alcoholic vapors, c) inacuracy in exactly attaining the identical temperature between initial and final solutions (which will cause differing thermal expansion and hence density), d) the effect of gases (such as carbon dioxide) which are more likely to be present in cider than in distilled water and perhaps e) chemical reactions in the boiling cider that will lead to density changes – a precipitate or sediment has been noted in boiled ciders and this will of course alter density.
On top of all of this the chemical composition and hence thermodynamic considerations for wine, cider and beer are different and hence this discussion is a very good example of what Andrew says repeatedly – all of these things must be calibrated for cider, not for beer or wine.
DuncanAndrew Lea
>
> The original method was described by a French chemist by the name of
> Tabarie in 1830. [Long history removed]
Tabarie is still nowadays cited as the originator of the 'extract'
method. You have done a wonderful job researching the past literature,
Duncan. I have only found transient references to most of that, and some
of it is very obscure and impossible to obtain outside of a specialist
library! Did you find all that in one inclusive source or did you have
to piece it together? BTW I am still pursuing a copy of Honneyman's
paper via the British Library but it is a tedious process. Hope it
doesn't prove to be a damp squib when I finally get it!
Andrew
Andrew Lea
> Andrew Lea wrote:
>> More significantly, acetic acid distills over and
>> I have been told that lactic acid can partially steam distill as well.
>> For this reason the determination of alcohol by distillation is usually
>> done on a solution previously made alkaline to neutralise the acids (see
>> Nick's note).
>
> Yea... I now remember reading somewhere that volatile acidity could
> affect distillation results and that acidity should be neutralised. I
> suppose this would be a more important factor in spanish ciders which
> contain more acetic/volatil acidity.
> Question - how much volatile acidity may a cider contain before being
> considered awful, and how could this amount of acidity affect the
> result of distillation (assuming distillation without neutralisation
> of acidity) and of Honneyman?
Quick 'back of envelope' calculation...... A cider with 0.2% acetic acid
might be on the verge of 'awful' but it is not a direct correlation
since most of what you detect is ethyl acetate. (Charley in 1932 wrote
... "It has been repeatedly observed that samples containing 0.25 - 0.3%
volatile acid may appear to the taste to be less acetic than ciders
containing considerably lower amounts). The SG of acetic acid is around
1.05. So, taking 0.2% as a yardstick, that gives an elevation of SG in
the distillate (made up to original volume) of 0001 degrees. For an
alcohol table at around 5.8% ABV (i.e. ca SG 9990 vs 9991) this
corresponds to a depression of about 0.07% ABV. Taking other minor acids
into consideration the potential error may be around 0.1% ABV depression
if the cider is not neutralised. I will leave you to do the Honneyman
calculation!
Zoecklein's "Wine Analysis and Production" also says that SO2 at 200 pmm
can depress the ABV value in the distillate by 0.2 - 0.5% v/v (because
SO2 is steam volatile from acid solution), likewise any wine in excess
of 0.1% acetic acid. Hence he recommends neutralisation prior to alcohol
determination.
For 'extract' distillation (relevant to the Honneyman procedure)
Zoecklein also notes "water, lactic and acetic acids, ethylene glycol
and 2,3 butylene glycol may be differentially lost during distillation.
Furthermore, dehydration products resulting from heating of sugars may
form, changing the density of the solution formed upon dilution back to
the original volume".
Andrew
Duncan
Much of the information came from abstracts published in the analyst in the 1890's . The royal society of chemistry has them in facsimile on the web but it requires an institutional log in unfortunately. The most recent paper I cited provided me with the clue that tabarie had developed the technique and subsequent searches of his name unravelled the debate regarding subtraction and division. Happy to provide the ten or so references but I'm not sure whether the pdf's I've downloaded are able to becirculated - they seem to cost £32 each!
"the analyst" also provided me with two nice articles on cider composition by barker and Russell, 1909, and "a contribution to the chemistry of cider" by Alfred Allen, 1901. Allen from memory was a subtractionist.
Duncan
Claude Jolicoeur
> Claude, my primary contention is that not only alcohol, but also solutes
> influence expansion and contraction of solutions such that these produce
> error in methods such as Honneymans.
button.... But I have been quite busy all day and didn't find the time
to correct it... And I sort of forgot about it also! until I read your
post...
But I think we need to put all this into perspective. No one ever
claimed this was an exact method, and I don't think any formula or
modification can make it exact. What we know is that the lower the
final density of the cider, the more accurate the method will be -
simply because if the finished cider has a low density, its properties
will be closer to that of a water-ethanol mixture because it contains
less soluble solids.
My feeling is that the method can estimate the ABV of a cider to +/-
0.5% (depending on the accuracy of the hydrometer used) quite reliably
(say 19 times out of 20 like in surveys) as long as the SG of the
finished cider does not exceed 1.025. For ice ciders with a SG of the
order of 1.060, it could be more problematic.
So I think this can be a very useful method to cross check the ABV
after having estimated it from the SG drop between juice to finished
cider. We know that the SG drop estimation is not exact because of
many factors (we have discussed this recently) and this estimation is
probably also accurate to +/- 0.5% ABV 19 times out of 20...
Then, if a cidermaker uses the 2 methods and gets very close results,
he can be quite confident on the result, as the 2 methods are fairly
independant. On the other hand, if the 2 estimates are very different,
the cidermaker may choose to send a sample of that particular cider to
a lab for analysis.
Claude
Duncan Galletly
But I think we need to put all this into perspective. No one ever
claimed this was an exact method, and I don't think any formula or
modification can make it exact. What we know is that the lower the
final density of the cider, the more accurate the method will be -
simply because if the finished cider has a low density, its properties
will be closer to that of a water-ethanol mixture because it contains
less soluble solids.
My feeling is that the method can estimate the ABV of a cider to +/-
0.5% (depending on the accuracy of the hydrometer used) quite reliably
(say 19 times out of 20 like in surveys) as long as the SG of the
finished cider does not exceed 1.025. For ice ciders with a SG of the
order of 1.060, it could be more problematic.
Then, if a cidermaker uses the 2 methods and gets very close results,
he can be quite confident on the result, as the 2 methods are fairly
independant. On the other hand, if the 2 estimates are very different,
the cidermaker may choose to send a sample of that particular cider to
a lab for analysis.
Duncan Galletly
As a postscript I wanted to ask Nick whether the lookup table used for
the official method was based upon cider data or ethanol-water
mixture. It seems odd that if the distillation method is the preferred
official technique yet might have an accuracy of +/- 0.5% then how can
it be used to regulate labelling that demands a tolerance less than
this. Very confusing.
Duncan
Claude Jolicoeur
reading are used See for example this one that has a range of 0 -
5%ABV for a cost of over 70€... In cidermaking, one would most likely
also need the 5-10% one.
http://www.brouwland.com/shop/product.asp?cfid=4&id=2080&cat=426&dt=24
Distillation method, done according to all guidelines with acid
neutralisation has a better accuracy than +/- 0.5; it is more like +/-
0.1% according to the claims I have seen. See
http://www.brouwland.com/shop/product.asp?cfid=4&id=373&cat=426&dt=24
You can then get full equipped fo distillation for about 570€
including 2 hydrometers. (Their ebullioneter is more expensive at over
700€).
Interestingly, I just noticed while searching for the above hydrometer
that Brouwland also sell an hydrometer (which includes a thermometer)
specially for the Honneyman (or a variation) for 20€. They don't give
much details however. They simply say that the second measurement is
taken after alcohol evaporation and that tables are provided.
http://www.brouwland.com/shop/product.asp?cfid=4&id=363&cat=426&dt=24
Claude
Nick Bradstock
The Spirits Table appears to be based on historical (and now old) data
collected by the (now defunct?) Laboratory of The Government Chemist
for/from the first distillation stage in the making of spirit drinks (whisky
- even whiskey, gin and vodka) made in the UK and specified in the various
Spirits Regulations. The reference appears to have been simply picked up
and transferred across to cider, wine and, I believe, beer made in the UK.
(Beer duty was, until 1993, calculated from wort SG before fermentation but
had to change to %ABV on the adoption of (what else?) EU Directive 92/83.)
As such the table would, I suppose, be based on fermented saccharified grain
mixtures - not fermented apple juice....
I guess we should have to ask the Queen's Remembrancer for confirmation!
Best
Nick
-----Original Message-----
From: cider-w...@googlegroups.com
[mailto:cider-w...@googlegroups.com] On Behalf Of Duncan Galletly
Sent: 04 August 2011 7:25 AM
To: Cider Workshop
Subject: [Cider Workshop] Re: Honneyman method for ABV
--
Claude Jolicoeur
> Interestingly, I just noticed while searching for the above hydrometer
> that Brouwland also sell an hydrometer (which includes a thermometer)
> specially for the Honneyman (or a variation) for 20€. They don't give
> much details however. They simply say that the second measurement is
> taken after alcohol evaporation and that tables are provided.
http://www.brouwland.com/shop/product.asp?cfid=4&id=363&cat=426&dt=24
VINOFERM with thermometer" - see above link.
They replied by sending a PDF of the user's guide for the product (in
French and another language which I think must be Flemish). From this
guide, the method described is identical with the Honneyman procedure,
and the table provided is almost identical (slight differences on the
second decimal, somewhat larger difference for very high %ABV). The
hydrometer provided is graduated in Oechsle (which are same as degrees
of SG) and appears to have a range of -10 to 40 Oe (i.e. SG from 0.990
to 1.040) with graduations at 0.5.
They also provide a table giving the amount of extract in grams per
litre from the second gravity measure.
Claude