Lychee-Guava Wine (using pulped guava juice)

Winemaking Talk - Winemaking Forum

Help Support Winemaking Talk - Winemaking Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.


May 25, 2009
Reaction score
I have a question about reliably measuring starting SG of a must that is a combination of a guava fruit juice with 30% very finely pulped fruit (a lot of suspended fibre so quite thick and gluggy !) and a canned fruit base (Admiral lychees in syrup).

I have decided to risk attempting my own recipe this time, since my last 4 batches of wine were all follow-the-recipe. As a kid I grew up with with four Rambutan trees in our back yard (a very similar fruit to lychees) and numerous Guava trees ... so I am about to start a Lychee-Guava wine from a canned fruit-in-syrup base (lychees) and a pulped fruit juice base (guava). I already have a yeast starter underway.

From Safeway I bought 3 litres of Golden Circle Guava nectar (30% pulped fruit; no artificial colours/flavours/preservatives). I also bought 11 cans (565g each) of Admiral brand Lychees-in-syrup (45% lychees).

Luc, after reading your recent comments on Lychee wine plus your blog recipe for lychee wine, and particularly after opening and tasting a can of lychees-in-syrup this morning, I too have to say that lychees would make an exquisite sweet white wine. So without wanting to overpower that delicate sweet lychee flavour, I am thinking of running with ...

a) 11 cans lychees x 565g/can = 6215g (lychees in syrup). Cans are 45% lychees so that equates to ~2797g lychee fruit.


b) 3 litres pulped guava juice at 30% pulped fruit content = 900g guava.

So I will end up with a strictly fruit ratio of Lychee:Guava of 3:1
(excluding syrup and additives).

The must will comprise 3 litres guava juice + ~5 litres lychee juice (i.e. syrup + juice extracted from all lychees during fermentation), equalling 8 litres total, so I will add 3 litres water to yield 11 litres of fermentable liquid.
(Enough for 2 x 5 litre demijohns with a little to spare for topping up).

I have a couple questions:

1) Will this amount of fruit and juice provide reasonable body ?

2) How reliable is my SG measurement of this must going to be to estimate inherent sugar, given that the 3 litres of guava juice that I add to the must is actually finely suspended pulped fruit which will increase the SG in its own right.

I am worried about getting the sugar addition wrong. i.e. if the SG is high "because of finely suspended guava pulp", I may end up adding too little sugar and end up with a wine that is too weak in ALC/vol. I should aim for around 11% ALC, but don't want to end up with 7-9% which would be too low for safe preservation of this wine.

Can I safely assume that once everything is blended together in the must, that the SG will reflect the sugar content to a close enough approximation, despite the uncertainties with the the very finely suspended guava fruit pulp.

Maybe I should independantly estimate sugar using the bottle+can ingredient labels, and see how this compares with the SG-indicated sugar ?

Another quick question - I assume that because these are packaged juice and canned fruit (thus presumably pasteurized), then it is not necesary to add a Campden tablet 24 hours before pitching the yeast starter, or should I add theCampden tablet just to be doubly sure ?

btw - the starter (lychee + guava) smells amazing.
I never did guava or a blend with it, but Lychees are definitely worth the effort. Lychee wine is just great !!!
I still have about 10 cans in my garage waiting to become wine, if I find the time.....

Now about the SG measuring.

You should mash everything up (except the pits) and
put pectic enzymes over the pulp, and add some sulphite
to prevent spoiling. Mix thoroughly.

And do not go Dutch. Hey I am a Dutchman so I can say that !!!
Use about a teaspoon pectic enzyme per kilo fruit.
Trust me it is worth it.
Pectic enzymes when used at room-temperature (24-30 degrees celsius) will break down even carrots (trust me I've done that).

Now wait for 24 hours and during the waiting mix regularly.
After 24 hours the pulp will have become mostly fluid.
The waiting is important because by soaking this long the
sugars, acid and flavors will get into the juice.

Now take a sample out and filter it through a coffee filter.
Take an SG reading from that.

It will be accurate.

Ahhhh I see ! Thanks Luc.

That makes a lot of sence. I now have the lychees, guava, extra water, petinase (enzyme) and 3 Camden tablets combined, and will stir every couple hours for the next 24 before adding my yeast starter.

After adding the yeast starter in 24 hours I will calculate SG and then add required sugar + yeast nutrient + acid (I hope my titration kit arrives in the mail tomorrow !), though I can add any required acid a few days into the ferment with no problem I guess.

Luc - I noticed your recipe did not use tannin - what made you decide to use just acid and no tannin. I guess tannin is just tannic acid anyway, another form of acid, ja?

Will post a pic of the primary in my album tomorrow. It has a nice pink-orange tinge from the guava but smells dominantly of lychee.

As a novelty I have decided to do the primary ferment in one of my new glass 15 litre demijohns (11 litre batch in a 15 litre primary) instead of the opaque plastic primary, so I can watch the rise and fall of the lees during the ferment which is fascinating, I could watch that for hours :D
Measure SG BEFORE adding the starter.

When fermenttion starts hydrometer readings are influenced
by alcohol that is formed. So the readings will be off a bit.
So always do the readings and additions to the must before
fermentation starts.

Wether to add tannin or not is a matter of personal
preference. Tanin gives some astringency, a kind of 'bite'
to the wine.
I do not always find the need for it in fruit wines.

Luc - I made 2 independent estimates of sugar content in my lychee-guava must before adding the yeast starter, and they agree reasonably well !

Method 1 – Calculating Sugar Requirement from SG (needs additional 759g).

I measured SG before adding the starter, but after the pectinase had been active for 20 hours (a reasonable time for the pectinase to release sugars and acids from the guava pulp and blended lychee fragments). Your coffee filter idea worked great in cleaning up the suspended pulp fragments and ensured the hydrometer worked smoothly.

SG was 1048. On my hydrometer that equates to ~133g sugar/litre.
11 litres liquid component of must x 133g = 1463g sugar in must (inherent from the cans and juice).

According to my hydrometer, to yield 12% ALC I need 202g sugar/litre. So, in total the must needs to end up with 11 litres x 202g = 2222 grams sugar.

The total sugar to be added is Required Sugar (2222g) subtract Inherent Sugar (1463g) = 759g sugar.

Method 2 – Calculating Sugar Requirement from Label data - (needs additional 693g).

Lychee Can Label - Nutrition Information: 19.1g sugar/100g, so 11 cans x 565g (individual can content weight) x 0.191 = 1187g sugar.
Guava Container Label - Nutrition Information: 11.4g sugar/100ml, so 3 litres x 114g = 342g sugar.

Thus total sugar coming from the canned ingredients is 1187g + 342g = 1529g sugar.

So total sugar to be added is 2222g less 1529g = 693g sugar (compare with 759g estimated using the SG method above, reasonably close).

Another way of looking at it, the 1529g of sugar derived from the label information is diluted through 11 litres of must, so the average amount of sugar per litre is 1529/11 = 139g. This should equate to a “predicted” SG of 1051 according to the scale readings on my hydrometer.

Thus while the “actual” hydrometer reading of the must is 1048, the “predicted” hydrometer reading based solely on sugar content listed on the labels would be around 1051 (a pretty good reconciliation !!). In fact, if the pectinase continued to act for another couple days and extracted most of the remaining residual sugar from the pulp, the measured SG of 1048 would expect to increase slightly further towards 1051, further improving the reconciliation between measured SG and SG expected from the labels.

What I did

OK – so I added 725g sugar (i.e. the mean value of the 2 independent required sugar estimates, being 759g and 693g).
Then stirred thoroughly and remeasured the SG of the must. This time the SG came out as 1066 (=10.6% ALC), so I added another cup of sugar (196g) which increased the measured SG to 1070 (=11.3% ALC).

I expect there is still a little sugar left in the mashed lychee pulp, despite the enzyme being active for a day, so I think this wine will come out a little higher than the 11.3% ALC, and closer to the hoped for 12% ALC.


My titration kit arrived just in time in the mail. It is much more accurate than those &%#$ paper strips ! The titration route is easy to use I found and gives me much more confidence that I have the right answer.

My must had total acidity (i.e. free + fixed acid) of 0.3% (quite low).
The instructions say white wine – aim for 0.7-0.8% total acids and for desert wines aim for 0.5-0.75% total acids. So I decided to adjust the must to an acid content of 0.7%.

Going from 0.3% to 0.7% I need to add 4g acid blend / litre, thus 44g for 11 litres.
Yesterday I added 10g citric, 10g malic and 10g tartaric (total 30g). Seemed like a lot weighing it out, that's why I only added 30g (I used my new pair of digital scales), but in reality, it's hard to tell when tasting the must. So tonight I will add another 7g malic and 7g tartaric to make up the 44g.

Like you, I might skip the tannin also.

Luc – I am happy with how this went, it has been a good learning experience using multiple ingredients in a self-adjusted recipe. Thanks ! :b

I have another question tho …..
You stated in your lychee wine blog that 18g sugar should bring 1% ALC to 1 litre of wine. My hydrometer scale puts the figure closer to 16.8g. I tried to read your link, but the explanation is double dutch to me (excuse the pun!) – do you have an English translation for that post? at

Looks like you averaged the conversion rate for Sugar-to-ALC from a number of sources listed in your table, to yield a more reliable average of just over 18g, however that is a little different to the commercial hydrometer I use which says 16.8g ?

I think my must mixture is close to OK now, as I can hear the yeasts (Vintners Harvest – Premium Wine Yeast SN9) happily feeding on lychee and belching CO2.

Last edited:
I would love to know what hydrometer you are using.

I have a story planned on my hydrometer which has 3 scales.
Alcohol, sugar and potential alcohol.

The last two where way off, so I had an e-mail exchange with the manufacturer and believe me the answer they gave me was hilarious.

Look at it this way.

Take 50 gram sugar and dissolve that in 1 liter water.
Now look at the SG AND look at the sugar scale.

Next take 100 gram sugar and dissolve that in 1 liter water.
Again look at both scales.

Do this again at 200 gram and look at both scales.

My own SG table (downloadable from my web-log estimates indeed 18 gram sugar for 1% alcohol.

So 200 gram would gain about 11% alcohol.
Your scale says it will give about 12% alcohol.

hhmmmm - I will measure out 50, 100 and 200g in 1 litre later tonight and see what my
hydrometer says !

Luc - it is a ... "Stevenson Reeves Ltd" hydrometer.
Made in Scotland. #S1011.

It has 3 scales:

1) SG at 20 degs;
2) Potential % ALC;
3) Approx! sugar grammes/litre.

(It appears at the top of the list when you google it).
Last edited:
Luc – My test results are very interesting at this early stage. I ran a test of my Stevenson Reeves hydrometer using fairly controlled conditions as you suggested above. (btw - I am presuming here that there is not too much variation in the mass of different types of stock-standard white processed cane sugar):

a) I used an electronic scale accurate to 0.01g to weigh the sugar.
b) I used a thermometer inserted into the hydrometer solution to ensure I took the reading only when the temperature was 19-21 degrees Celcius. The hydrometer is rated for liquids at 20 degrees Celcius.
c) I read the hydrometer by looking at an angle from just below the surface, to ensure I was capturing the base of the meniscus (not the top edge).

These are the results:

Actual 50g sugar in 1 litre water: Hydrometer reads 60g/litre ! (20% error) i.e. 10/50
Actual 100g sugar in 1 litre water: Hydrometer reads 107g/litre (7% error) i.e. 7/100
Actual 200g sugar in 1 litre water: Hydrometer reads 206g/litre (3% error) i.e. 6/200

Luc – I think it is very interesting that the discrepancy (% error listed above) varies in a systematic manner, (i.e. the error decreases to just under half of the previous error with every doubling of the sugar content from 50 > 100 > 200 ! That tells me something …

It would appear that the “Approximate Sugar grammes per litre” scale has slid up by the equivalent of about (on average) 7.7 units or 7.7g/litre. That would result in higher relative % discrepancies at lower sugar contents and lower relative % discrepancies at higher sugar contents.

I will repeat this over a number of days, with some repeats at 50, 100, 200g and also some at 25, 75, 125 and 175g and then plot and run a regression and so establish my own scale that is tuned to my brand of sugar.

I am away in Indonesia as of Monday next week for work (busy preparing for that now), so I will do these follow-up data-points when I return and post a graph.
Last edited:
I have another question tho …..
You stated in your lychee wine blog that 18g sugar should bring 1% ALC to 1 litre of wine. My hydrometer scale puts the figure closer to 16.8g. I tried to read your link, but the explanation is double dutch to me (excuse the pun!) – do you have an English translation for that post? at

Looks like you averaged the conversion rate for Sugar-to-ALC from a number of sources listed in your table, to yield a more reliable average of just over 18g, however that is a little different to the commercial hydrometer I use which says 16.8g ?


What I did was I took all my Dutch books and examined the SG tables in it.
Next I took from each book several readings at fixed SG points and made an average of that. And I calculated the overall average of all writers.
As you can see in the table the figures they all give have a lot of variation.

Now I looked at websites from several yeast manufacturers and found that only one gave a figure on how much sugar was consumed by their yeast to produce 1% alcohol.

So as I do not have access to a real laboratory and have to do all my tests on my kitchen-table laboratory I had to work by estimates and available literature from the 'great winemakers' we once had.

Lately I have ordered by mail-order some US books.
I have Iverson's one and Pambianchi's one.

Iverson gives at an SG of 1080 a figure of 21 brix (page 214 of Home winemaking step by step).
At page 161 he states that 55% will be converted to alcohol. Now that would be 11.55% alcohol.
Giving that 21 brix is about 230 gram alcohol sets Iverson in the 19.91 gram league !!!

Pambianchi at the other hand gives an SG of 1080 a brix value of 19.4 and converts that to 10% alcohol (Techniques in home winemaking page 465).
Pambianchi however says that will be 20.9% sugar. So in a liter that is 209 gram sugar.
So 209 gram giving 10% sets Pambianchi at even 209/10 = 20.9 gram sugar per 1% alcohol !!!!

Jack Keller gives at his homepage an SG of 1080 an alcohol percentage of 10.9%.
Not really far off from my figure which says 10.4%

So both Iverson and Pambianchi estimate a higher weight off sugar for making 1% alcohol as I do.

Remember there are great math problems here involved.
Presume you add a certain amount of sugar to a must then consider this.
- Some of the alcohol will be blown off through the airlock or evaporate.
- 1 kilo sugar does not have a volume of 1 liter. No it is just 600 ml.
- When this volume is converted to alcohol will it expand and how much.
- about half will be converted to alcohol. But is it really 55% or is it 58% or just 53%
- When alcohol and water mix they will contract, meaning that the total volume will be less as the volume of the two seperate fluids.

There are very complicated processes going on. Therefore I do believe in all honesty that my calculations are not far off and you can use my SG table without fearing to be way out of line.


I agree 200% with your posted comment ..... "Lychee wine is just great !!! "

This batch (mostly canned lychee with a little guava juice) is now almost 9 months, and I have to say, it is luscious, sweet, smooth, well rounded with great body (I am talking wine). When chilled - it makes a genuinely superb wine.

Out of the 7 different wines I have made (5 tasted), this will be one of our quaffing regulars I can see :dg so will need to get another bigger batch underway soon

Latest posts