The Taleggio Experiment

The decision to rent the School of Artisan Food was made, a recipe was researched, whilst attending a course there, I was able to work out what equipment we’d need and finally our dates rolled around.  Rose purchased containers, filled them with milk and sent them off with a refrigerated courier.  I drove myself off to the School of Artisan Food and got ready to receive milk.
I spent a day sanitising equipment and writing up a HACCP plan for our trial production and the following day was in bright and early for the milk to arrive.
Taleggio is an interesting make from my point of view, in that it uses thermophilic bacteria as a starter culture and yet doesn’t use the temperatures at which thermophilic bacteria tend to work best.  Like all washed rinds, the curd doesn’t want to acidify very much and it wants to retain a calcium rich, pliant structure.  The thermophilic bacteria therefore are used precisely because they will start working but as the temperature of the make cools off when the cheeses are in their moulds, they will stop going, the acidity will level out and won’t develop further.  This in theory and coupled with curd washing, should mean that the cheeses remain pliant with their moisture locked within the curd structure and soften when ripe to a gloriously oozing texture.  That is the theory anyway.
Taleggio photo courtesy of the
Within this, of course, there are many parameters to play with.  So many, in fact, that I wish we were in full production right now in some ways so I could be happily making cheese day after day, tinkering with a whole multitude of variables.  Would a degree or 2 more or less in terms of temperature affect the rennet set and the texture as the cheese matures?  How would the flavour and acidity be affected if I remove a bit less whey at curd washing?  What if I add in more starter cultures at stirring?  What if I stir for longer?  That’s not even getting started on how much starter we need to use to work with our milk and how much rennet will get me a 15 minute flocculation and 45 minute hardening time (which, I believe, is what I’m aiming for).
First challenge and challenge not yet overcome at that, is the quantity of starter.  There are no hard and fast rules for this of course because the amount of starter you use is entirely related to the numbers of lactic acid bacteria in your milk.  Thinking back to my Holker Farm days and remembering the drainage battles we had balancing acidity and calcium, I figured that if I wanted to have a slow acid development, even though I was using thermophilic bacteria this time rather than mesophilic ones, I should be using pretty small quantities of starter.  In retrospect, I’m not sure that was the case, but you live and learn.  I remembered that when I left Holker, we were using tiny quantities of bulk starter, having been advised to drop to around 0.025% and before that had been using still pretty tiny quantities at 0.05%.  I decided to start at the higher of these values, having made up a yoghurt culture in skimmed milk the night before and incubated it overnight.  Yoghurt cultures, for those who didn’t realise, are thermophilic bacteria.
Now, at this stage, the benefit of recording values of acidity began to hit home.  In all my time at Holker, we never recorded a pH.  The pH meter had broken before I arrived and they are very expensive bits of kit to replace.  We took titratable acidity of course but the TA of our starter cultures which I took every so often, are hard to correlate with that of this yoghurt because of the buffering factor.  Our starters at Holker were made up in sheeps milk, which is high in protein – it can be up to 3 times that of cows milk.  My yoghurt cultures were made up in UHT skimmed cows milk.  As you all no doubt remember from the pH and Titratable acidity post last autumn, protein captures Hydroxl ions (OH-) when you add the alkaline solution looking for a pink colour produced by its reaction with the indicator, phenolphthalein.  This distorts the correlation between acid and pH because it is non-uniform.  The more protein the milk has, the more Hydroxyl ions it can capture and the more Sodium Hydroxide needs to be added before a reaction with the indicator will register.  In other words the TA value will be higher in sheeps milk than in cows milk just down to the protein.  In fact, at Holker when we began making cows milk and sheeps milk cheese side by side, we noticed a huge difference when recording the TA at 24hours (or thereabouts) between the two.  Our sheeps milk St James regularly recorded 80-90 ’D while the cows milk Brother Davids struggled to reach 40’D.  You would think that the majority of the protein in the sheeps milk had been locked up in the curd by then but just as the milk is higher in protein, so is the whey and so the titratable acidities ended up being quite dramatically different.
Anyway, returning to the matter in hand and hope I didn’t lose too many non techno cheese geeks along the way.  There was no point, to my mind, trying to correlate vaguely remembered TA values of starter culture with a yoghurt I had just made as I didn’t have any of those values recorded for reference.  So I took a pH reading of the starter and was conscious it was more acidic than my notes from Ivan Larcher’s course suggested was ideal (pH 5 – to make sure you catch the bacteria while they are multiplying happily and before the lactic acid they have produced can denature them and kill them off), but otherwise didn’t have much to relate it to.
On the first trial, 0.05% in quantity was added to 50 litres milk, the milk was heated to 34’C, rennet added at the appropriate pH change and I filled my pot of water to look out for flocculation times.  The flocculation happened right on time, the hardening more or less followed the pattern it was supposed to.  I pre-cut, then cut to hazelnut size (more or less – it’s a bit hard to use a cutting harp designed for a big vat in a 50 litre tub), let if settle, took off the 25% whey, added back the appropriate quantity of water at 32C, added some starter back in for flavour and stirred.  The recipe was one I’d found online and frankly has already been adapted.  At the time, I queried curd washing with Ivan Larcher and he replied
‘Good luck settling the pH at 5.2 without it.’
Later on, I asked for clarification on quantities and adjusting those parameters only to be told that it’s an industrial technique and he didn’t recommend me doing it.  I have therefore stopped.
However I was doing so on this trial make, and at every stage, I was recording pH on the spreadsheet Ivan had emailed me after our Blue Cheese course so that it would track the pH curve.  Unfortunately my pH curve didn’t curve.  It was more of a wobbly straight line.   Short of leaving the whole thing for 24 hours to acidify on its own, there wasn’t much I could do but proceed, pre-draining the curd on a mat and then filling the moulds with the drained curd pieces and turning, turning, turning throughout the afternoon.
All looking quite convincing so far – unfortunately it’s all in the maturation.
They looked pretty convincingly like cheeses.  They were draining.  But who knows what was going on below the surface without much acidification.  The problem is, all sorts of other bacteria could be enjoying the quantities of lactose and developing to undesirable results.  Unfortunately despite doubling the starter cultures the following day, the same acidification pattern followed.  Evidently at Holker Farm the starter culture had very minimal effect and acidification was largely governed by the lactic acid bacteria naturally present in the sheep and cows milk.  Further unfortunately, I already knew from a lactofermentation we had done a week or so earlier when Rose drove up (bringing a bottle of milk with her) to SAF to meet Ivan and me after one of the days of the course had finished, that we didn’t have a lot of strong lactic acid bacteria in our milk at the moment and that other things tended to become dominant.  To say I was nervous of the test results we would get from milk and curd samples would be an understatement.
I left SAF; the samples headed to the lab; we waited for the results.  As I had feared, without enough lactic acid bacteria from the starter or naturally present in the milk itself, enteros and pseudomonas had had a field day.  Staph. aureus hadn’t done so badly either.
Not quite what I was hoping for but still looking relatively like cheeses
After a couple of weeks, I drove to SAF to collect the unfortunate cheeses.  I did not have high hopes to be honest, particularly as one of them had pancaked overnight and collapsed – a bit of a surprise for me and also for Lee Anna.
However, I had known, that I had more cheese than I was expecting, which I suspected meant they were too moist.  This raises the likelihood of crazy things happening during maturation.  What I didn’t know however until I began to think and mull it over was that those rather too healthy pseudomonas might also have played a part in this too.  Pseudomonas, as I had discovered thanks to the very knowledgeable Paul Thomas, are caseolytic (they eat casein).  Could that mean that they might increase the speed and amount of protein breakdown in our cheeses?  One quick email and a reply later and yes, by no means the only factor but, if there were large numbers of pseudomonas, then there was much more chance of pancaked gooey cheeses that fall apart.
All in all, I wasn’t sure what I would find at SAF but although one batch had fallen to pieces, the other did seem to be holding some shape and smelled convincingly washed rind.  Not the best behaved of washed rind, I’ll admit, but I’ve smelled worse in my time.  We tasted one of them and to my surprise there wasn’t a strongly bitter flavour that I was expecting due to the pseudomonas, in fact the predominant flavours were beery, yeasty and fruity with a hint of meaty and savoury in the background and perhaps just a touch of the bitterness on the rind but certainly not overpowering.
Now let’s be clear, it’s not the flavour profile I want but then again, the recipe didn’t work, so for it to have turned out to be not only edible but while a bit raucous, actually not too bad, was a definite bonus.  That said, a valuable lesson was learned for Cheese Trial no 2: use a hell of a lot more starter!

Acidity and Cheesey Chemistry

Before any of my long standing friends wonder who I am and what I’ve done with Anne, I should point out that the chemical knowledge in this post is courtesy of the collective brains of Dr Jemima Cordle, Dr Katie Jewell and M. Ivan Larcher who have very kindly either taught me the info or have answered my questions on the subject.  I hope I have interpreted it correctly.  Any chemists out there, please pick me up on any mistakes so I can correct them.

The difference between titration and pH is that you use titration to calculate the concentration of Hydrogen ions by means of measuring the amount of alkali ions you have added to cause a certain colour change.
A pH measurement is the log (base10) of the concentration of Hydrogen ions.
pH meter at Neal’s Yard Creamery recording acidity development of Finn curd
Titratable Acidity kit set up at Old Hall Farm
Hydrogen Ions, Hydroxyl Ions and Acidity
In a basic solution there will be Hydrogen ions (H+) and Hydroxyl ions (OH-).  The H+ ions determine the acidity of a solution and the OH- ones are in predominance in a basic solution (when you stick both ions together you get H2O which is water)
Although pure water is present largely in its molecular form rather than atomic (which means that there are bonds linking the H atoms to the O atom to form the water molecule) a portion of it is ionised and present as H+ and OH-.
Because these two ions are present in equal quantities in a pure solution of water it remains neutral with a pH of 7. (The actual pH number is derived from the concentration of H+ present). 
When substances are dissolved in water they often disturb the H+ OH- balance. 
When the concentration of H+ is higher than the concentration of OH- the solution is said to be acidic and the pH is less than 7. 
When the concentration of H+ is less than the concentration of OH- the solution is said to be basic and the pH is more than 7. 
Acids and bases can vary hugely in strength. Hydrochloric acid for instance has a lot of H+ ions present and has a pH of 1. Citric acid on the other hand has less H+ ions present in it and has a pH of around 2.5. 
How the pH meter measures
A pH meter is able to directly measure the concentration of H+ ions present in a solution into which the probe is stuck. It then converts this concentration to the pH reading. 
How the TA measures
Titratable acidity measurement on the other hand is an indirect technique to measure the H+ concentration. There are certain ‘indicator’ solutions available that change colour when the solution they are in gets to a certain pH.
In cheesemaking we use phenolphthalein which turns pink when the solution gets to pH8.3. So we add a basic solution (NaOH or sodium hydroxide) to the milk or whey until enough OH- have been added to cause the colour change.
The amount that has been needed will tell you how many H+ ions were present in the original solution.
To milk you need to add very little NaOH because milk contains a very small amount of H+ but to whey from cheese the morning after making, you need to add a lot more NaOH because the lactic acid bacteria have produced a lot of acid and therefore a lot of H+ ions are present.
Having said this there is a complication which means that the pH/titratable acidity measurements are not 100% correlated. The complication is an interesting acid/base phenomenon known as buffering.
Buffering happens when an acid or base is added to a solution, but the pH does not change. This happens in cheesemaking because there is a lot of protein present.
The protein is able to capture H+ ions as they are formed and therefore prevents them from ‘being seen’ in the solution. They do not add to the H+ concentration that determines the acidity.
This can also happen with OH- ions – the protein can capture them and stop them from causing a pH change.  
The latter is relevant during titratable acidity measurements because sometimes even though OH- ions are being added they are not being used to neutralise the H+ present and raise the pH to 8.3.  Instead the protein is capturing them.  This distorts the relationship between pH and TA in a non-uniform way.
It will depend on the specific composition of the milk and the stage of acidity that has been reached. 
Comparisons of pH and TA therefore on a daily basis can help you extrapolate some physical properties of the milk you are working with – see the chart that follows.


Making Starter Culture

This week we were down to our last pots of frozen starter and needed to culture some more ourselves. We had done a trial run about 4 weeks ago which seemed to have worked so on Monday & Tuesday we made up another couple of batches.  Monday’s was made using St James whey and Tuesdays’s was made using milk.  As yet I’m not really sure what the difference will be between the 2 of them or indeed if there will be any.  One theory says that the starter will work more powerfully in a less fat rich medium than milk as the fats can inhibit starters whereas in whey, there is available lactose for them to consume but hardly any fat.Some cheesemakers culture up starter every 2 or 3 days and don’t freeze it, but for the quantities that Martin needs to use it in, this is simpler and if it’s frozen at the right time you don’t harm the bacteria.  They need to be put in to freeze in the lag phase before they start reproducing because once they start to reproduce they’re vulnerable.

So on Monday we made the starter up in whey (largely to be honest because I’d already set both vats of milk by the time Martin came down to the dairy).  And on Tuesday I took out a  litre of milk from one of the vats before I added the starter or rennet and we made up starter in milk.  The process was basically the same though – we poured 1 litre of whey into a sterilised stainless steel container which was then put into a pan of water and put on the hob.  The water in the pan boils but the whey doesn’t reach the same temperature and is basically pasteurised.  It then needs to cool down of course because the bacteria in the starter work best at around 30C and will be killed off at the sort of temperatures involved in pasteurisation.  Interestingly when heating whey, of course, you make ricotta, so the curds need to be strained out after it has cooled and before the starter can be added.  This doesn’t happen when making milk starter but you do have to dispose of the skin that forms on the top of it as it cools.

Each day we take a pot containing about 100ml of starter and in general we tend to use only about 65ml at the most so there is some spare. Once the whey had cooled down, therefore, we added 10ml of this starter to the whey and stirred it around to distribute.  If making in larger quantities, the rule would be 1% starter.  Then the whey and starter was poured into pots which were chilled in the cold store for about 20 minutes before going into the freezer.

To use, we take them out of the freezer the day before at ideally about 12 or 12.30 and it takes about half an hour to thaw out and then the bacteria get to work.  By the following morning when we’re ready to use them, they will have acidified.  If using the milk starter, as it acidifies it will also thicken to look less like a pot of milk and more like yoghurt.  The whey starter however doesn’t so we have to do a titratable acidity test on it to check it.

So far I’ve only used the new starters twice though so it’s hard to draw any meaningful conclusions on which is more active and effective from 2 TA readings but it will be interesting to see if the theory that the fats in milk inhibit the starter does actually work out in reality.

Monday’s starter made in whey
 Tuesday’s starter made in milk