Author Topic: Project Complete…2y  (Read 3798 times)

0 Members and 1 Guest are viewing this topic.

Offline Icelandr

  • Supporting Member
  • *
  • Posts: 1490
  • Age: 73
  • Location: Gabriola Island, British Columbia
  • Infinite Monkey Theorem at work
Re: Project Complete…2y
« Reply #40 on: December 30, 2019, 10:53:46 PM »
Steph, thank you for your response, I respect and admire your pursuit as different as it is to mine. You have shown a passion, obviously different than mine, but a commitment to learning more and improving as you go. I am, I guess, random abstract, and enjoy life through that filter. I recall one professional day in teaching where we filled out some kind of questionnaire, then broke into groups based on the “results” . . . .looking at the group I was in, abstract random, looking at one of the other groups, Concrete Sequential, I looked back and thought WOW, I don’t hang out with anyone in that group, we are so different. I am old enough now to see the benefits of diversification, but to this day I know deep inside, I would not last in that camp. Not better or worse just different.
Our goals are similar . . .make the best pizza possible, keep trying, knowing that somehow satisfying the definition in your head will never come, and in my case based on the last 70 odd years, find a new challenge.
Happy New Year!

PizzaParty 70x70, saputo floor

Offline stef

  • Registered User
  • Posts: 33
Re: Project Complete…2y
« Reply #41 on: January 02, 2020, 01:03:05 PM »
Part fourteen of the Pizza Project:


If you can’t measure it you can’t repeat it.     ....but rules are made to be broken.

Below are details on how measurements were made to obtain the project data presented in previous posts.

It may be useful for those interested in straightforward accurate home measurement, also for those who may be a little dubious about some of the numbers presented in earlier posts.

Much of the dialogue below is about calibration.  What does this mean to us ? 

It is checking the readings of your thermometer or weighing scale against a known accurate reference, so you find out if the readings you get on a day to day basis are actually accurate, or often more important, reliable: do they they mean what they say  today and tomorrow ?

Simplest calibration example is putting a thermometer probe into water which has a “very bubbly” rolling boil: it should read 100C, 212F.  A weight example is weighing a large number of currency cents, dimes, nickels etc: they are of known Treasury weight (approximate) but if you measure say 10 – 100 coins in one go and they are shiny, the average weight is likely to be correct.

Once you know if your thermometer or scale is giving an incorrect reading it’s no problem: you allow for the error when using it next time.  If you have a series of intermediate known temperatures or weights, you can check again and find out if the error is consistent, or changes with bigger or lower values.  And so on.

Does this really matter ?  Yes it does especially if you are comparing with somebody else's numbers.


Accurate temperature measurement is essential to manage pizza variability. 
Most of this pizza project has been completely dependent upon consistent accurate temperature measurement and control.

My reference thermometer (against which all others are compared) is a Thermapen 4.  This is manufacturer calibrated to an accuracy of +/- 0.4C (+/- 0.7F), it displays to 0.1C (0.2F).  It also quickly adjusts to temperature change.

The other thermometers used have been calibrated against this in the range of 10C to 30C (50 - 86F). 

The Thermapen is used for my ingredient and dough temperature measurement.  Its fast measurement speed is especially useful when measuring dough ball temperature at the point of tipping out of the tub onto the granito surface and checking FDT.

The cool box temperature is monitored with the thermometer sensor attached to the Inkbird temperature controller, which is calibrated separately to the controller function.  The temperature reading is the same as the Thermapen to 0.1C, but it is very slow in response, perhaps as part of the controller logic to reduce rapid heater / cooler switching.  The sensor is bonded to an aluminium plate and placed in the cool box between the second and third dough ball tub, in the tub stack of four. 

A ThermoPro TP-08S dual probe remote thermometer intended for BBQ monitoring is used for remote monitoring, monitoring of the other thermometers and as a fallback.  Compared to the Thermapen this is surprisingly accurate, but slower in response and reads only to 1C display resolution.   It works remotely at distances up to 20m, but does not work when there are substantial walls between the transmitter and receiver.  When it is used in the cool box one sensor is placed between the first and second tub, and the other between the third and fourth tub, to give a total of three readings up the stack of the tubs.

The above thermometers were calibrated together at the same time, the sensors were placed immediately adjacent to each other and bound together with electrical tape, placed in a small plastic bottle so that the bottle air would act as a buffer (ie average out air current / EMR induced differences).  A cup containing cold, then hot water was positioned next to the bottle, moved closer and further away to give the calibration temperature intervals needed.  See attached picture of this, towards the bottom right hand corner: please excuse the “unusual” looking arrangement, and no it was not Christmas.  There was no  hysteresis when the temperature was increased and then decreased (this means reducing temperatures change by the same amount as when as increasing temperatures).

All in all, surprisingly consistent out of the box: the low cost tools work well.

I also have used a $4 digital thermometer (accurate but so painfully slow) and an analogue milk thermometer (fast and good to 1C, if you use a magnifying glass to read the scale) in the past.

However, you do all of the above and unless you measure the right part, of the right thing, in the right way the result gives you a false sense of accuracy.  Like measuring an air temperature rather than a dough temperature etc: beware.


Accurate weight measurement is essential to manage dough variability: The project needs the same results next week and next month.  If I want to play around with the dough, fine: but not when guests are due. 

The world of weighing has become dramatically different in recent years, as compared with some time ago.

Importantly, weighing scales are now available at very low cost: 40 years ago weighing accuracy cost a LOT of money, and was only available to labs and specialist operations.  Now accuracy is available to everybody: at a similar cost to those archaic cup and spoon measures. 

However.... good practice around weighing ingredients has NOT changed.  You still have to do it right to get the right result.

Two scales are used in the project: one "kitchen" type for flours and water, and a mini scale for small weights, particularly yeast but also salt.  If I used oil, but I don’t, the mini scale would be used for that too. 

Both scales are digital display strain gauge load cell devices, with built in single chip converter and controller.  Cheap, generic technology mass manufactured with astonishing deals available everywhere.  Different in capability to laboratory and process industry type devices however... but astonishingly accurate.

Even so, it is essential to follow guidelines from the professionals, mainly applicable to the small weights.  Some of the key things that were drilled into me are below...

Use the "Tare" function frequently, also switch on and off frequently to avoid scale weight creep.  NO room draughts, extractor or a/c fans, breezes or breathing on the scale allowed: work in a corner and even use the small scale cover (I don’t).  Use the scale on a heavy stable platform (for instance, a stone work surface, or concrete floor) to minimise vibration. 

Locate away from electrical / electronic devices to minimise EMR effects on the scale controller electronics (especially as they are cheap and unshielded: don’t use the scale next to a cell phone or electric motor for instance).  Frequently, but informally, recalibrate against a single weight as a spot check to avoid problems ( I use a 24ct gold earring that I don’t like, it’s 0.169g: if the scale shows 0.171g for instance I switch it off and on again, this usually works: the earring is kept with the scale so I don’t forget to use it).  Nevertheless temperature effects on the strain gauge sensitivity are beyond control as my temperatures vary a lot (temperature does not affect weight, temperature affects the weighing machine particularly the strain gauge and bridge).

The key pro lesson however is .... a scale should NEVER be relied upon for the lowest decimal place to which it can be read (order of magnitude).  This means if the scale can read to 3 decimal places (say 0.001g on a mini scale) only rely upon the result at the 0.01g level, an order of magnitude greater. 

For example a reading of 0.023g could be optimistically taken as 0.02g (round down), a reading of 0.028g taken as 0.03g (round up).  Pessimistically, ignore the last digit completely, so the weight is 0.02g.  I use changes in the last digit as a “warning” when weighing: that the next digit is about to go up: ie with yeast when the scale gets to say, 1.739g, a few more grains of IDY and it will be at 1.74g.

If the scale can read to 1g (a regular kitchen scale) only rely upon the result at the 10g level. ...Bet you won't.  So a reading of 536g should be taken as 540g (round up), a reading of 532g taken as 530g (round down).  Or 530g.  That is 1% different.  I find that a bit extreme, we’ll see when we get to the actual calibration.

If your regular kitchen scale reads to one decimal place (say 0.1g), It would be best not to trust that digit.  However, if you're not risk averse, calibrate it at the least (If you are on US Dollars, you can weigh 10+ shiny nickels, then 10+ shiny dimes: if the difference is correct, it is likely a reasonable scale at those weights.  Note there is variation between every single coin, so taking 10 gives a reasonable starting average for a regular scale in the lower weight range.)   

Not many people take notice of weighing scale limitations, believing that every scale digit must be right. Also known as paying for the bigger numbers.  However it is CRUCIAL with the mini scale for yeast weighing. 

Starting with the mini scale for small weights: mine is a “TL-series” generic Chinese mini scale for weights up to 50g (a bit under 2oz).  It does not register weights below 0.01g, the lower limit of scale sensitivity.  It comes with a 50g calibration weight, good for the limits of the scale but not helpful for intermediate weight accuracy.  The scale displays to 3 decimal places, which means it can display for example 49.999g.  There are many variants of this scale on Amazon and Ebay: you can see it in one of the attached pictures.

Calibration: to check for possible errors in the up to 50g range I used a set of uncertified commercial calibration weights in the range of 0.01g up to 50g.  They claim to be accurate to +/- 0.003g.  These were purchased because I cannot find any repeatable small weights for calibration: pharmaceutical tablets and currency coins are not consistently accurate enough.  (you may find a friendly jeweller or gold dealer who may allow you to bring your scale in so you can check it out)

The measurements were made at 22-24C in a draft free corner of the closed kitchen, the SAME location I use for all of my weighing for pizzas.  The scale was repeatedly tared, turned on and off throughout the calibration and different weights were used in combination to double check individual readings.  The scale was recalibrated twice during the measurement.  There was no hysteresis.  The smallest calibration weight, 0.01g, sometimes would not give a scale reading at all as it is close to the lower limit of scale sensitivity (around 0.01g) below which it does not register (ie the scale displays zero)

The calibration showed zero error in the 1 - 50g range, as shown on the attached graph.  The graph vertical axis is calibration error (Difference) in g, worst case error being in a range between 0.003g to -0.001g.  Note 0.003g is the claimed accuracy limit of the calibration weights.  Each measurement is for an INDIVIDUAL calibration weight, to avoid cumulative errors caused by using multiple weights together.  The x axis is calibration weight in g, but using a logarithmic scale to emphasise the low weight part of the calibration, and "compress" across the higher weights on the right hand side. 

I have used Difference g= (Scale reading g) - (Calibration weight g),

So if the difference is a positive number, the scale is reading high, if a negative number, the scale is reading low.

The WORST case errors would mean that 0.01g on the scale readout would actually be 0.011g. (a 10% error), and for 0.2g on the scale readout would actually be 0.197g (a 1.5% error)  Whilst -0.001g error sounds small, as a percentage of a very small weight it is significant.  As the weight gets bigger the error % will typically reduce proportionally.

Back to reality: how accurate is the scale over the ranges used for my pizza making ? 

The salt weights I used in the project are generally in the 10 to 20g range, primarily around 12 -16g.  Here there is no evident scale error.  So my regular salt weight of 12.4g will be accurate, bearing in mind real world variations in air humidity and foreign bodies in the fine sea salt (shell or sand) will also be at play.

Most important, yeast.  The amounts used in the project are typically between 0.05 and 0.1% IDY, which normally calculates out to 0.25g - 0.5g for my 4 dough ball mix.  I use a maximum of 1.5g for big mixes.

How does this affect my pizzas ?  Let’s take a 14h planned Fermentation Time at 24C, which is close to the lowest yeast amount used for the controlled project runs. 

14h fermentation requires 0.248g of IDY for my 4 pizzas.  When weighed on the scale, it could be 0.248g minus 0.003g, which is 0.245g.  That will not affect the Fermentation Time measurably.

The calibration graph goes down to weights an order of magnitude (a decimal place) less than what is used (0.01g), so I have not broken the pro rule set out above.

As a complete aside, a 48h fermentation at 17C could use very roughly 0.03% IDY, or 0.15g for my 4 dough balls.  So extended fermentation times at lower temperatures would involve yeast amounts which can be accurately weighed by the scale.  Since the project completed I have gone to 58h at 24C, using around 0.04g of yeast.  Still ok.  Well, the yeast amount at least.

However the scale error is SMALLER with weights above 1g, so I weigh my yeast inside a container weighing 15g which is "tared out": the scale is therefore working in the comfort zone of the scale strain gauge, and the lower sensitivity limit is well avoided.

I am now happy using the scale for yeast. 

For larger weights I use a 15kg Globaltronics scale quality tested by LGA in Germany (but not certified) with claimed accuracy to 1g.  It does not register weights below 2g, the lower limit of scale sensitivity. 

To calibrate this scale two approaches were taken: firstly using the calibration weights used with the mini scale for the lower weight range, and then the mini scale was used to weigh 11 decorative quartz garden pebbles each weighing less than 50g. (picture attached).  I used the pebbles and a 50g calibration weight to calibrate the Globaltronics scale with a larger weight range (picture attached).  Using various combinations of the pebbles and 50g weight, I calibrated up to 960g.  I then used weighed kitchen crockery (small plates and cereal bowls, each around 380g) to extend the measurement to 4kg.  Any error in individual pebble and crockery weights will accumulate using this method, possibly averaging out, but it should give an indication of error trends. 

Attached is a calibration curve for the scale using the calibration weights, up to 260g.  The x axis (scale readout) is logarithmic, to provide more emphasis on the lower weights.  Below 50g looks a bit bumpy, greater than 50g looking good up to 220g  There is a slight drop off above 220g, with the scale under-reading consistently by 2g above 240g (difficult to see on the graph).  The next curve is a linear scale ‘blow up’ of measurements less than 50g.  Below 45g it’s a bit variable: the drop off at the bottom is where the scale becomes insensitive to the calibration weigh: the limit of resolution. 

The next attached curves are calibration curves for the bigger weights (garden stones etc).  The left hand vertical axis shows the weight difference between scale reading and calibration weight as a % ( -2 to +2% ).   This is shown as the blue curve. The right hand vertical axis shows difference between scale reading and calibration weight in grams ( -2 to +2g ).  This is shown as the red curve.  The horizontal scale is weight (0 to 960g ).  The difference in g is described by the equation at the top shown as the green curve, slightly dropping away with higher weights.

This scale has substantial error for weights below 40g: confirming the calibration weight data above.  I would not use this scale for weights of less than 50g and do not know if the errors in this range are typical of this type of scale or not.

All three curves show a downward trend to the right, indicating that the scale has an slight increase in negative error with increasing weights.  This means the scale display reads less than the actual weight.  It again confirms the slight drop off above 220g using the calibration weights.  Above 300g the % error decreases, being extremely small.  Above 500g the g error becomes consistent around -1.2g, below this it is round -0.5g.  This calibration behaviour is maintained up to 4kg at least, with a consistent variance between adjacent readings of around 1%.  There was no hysteresis. 

How does this affect my pizzas ? 

Up to 50g only concerns my salt measurement: so I use the mini scale for salt. 

My next weighing zone is around 100 - 300g (water, 100g representing 30% of a blend of two different temperature waters to hit my FDT) and 50 to 500g (flours, 50g being a 10% flour blend component) for routine pizzas.  However I weigh the water in the mixer bowl which weighs 900g, and the flour is weighed in a plastic bowl which weighs 65g. So my pizza range of interest is 115 – 1200g, occasionally higher for guest events. 

Looking at the vertical scales on the graph: the error amounts are small - I have chosen to ignore them.  I could use the equation to compensate for error, but hey, I do have a life really.  With flour, I simply do not know if humidity and flour storage effects introduce more variability than the weighing scale.

Having gone through all of the calibration effort I feel my results are unlikely to be substantially in error due to equipment limitations. BUT of course human error is there: my mistakes... and I cannot control the variability of the ingredient properties.


ps Apologies for the mind numbing stuff.  Please do not use cups and spoons though.

Offline stef

  • Registered User
  • Posts: 33
Re: Project Complete…2y
« Reply #42 on: January 04, 2020, 02:47:35 PM »
Part fifteen of the Pizza Project:

e2e Process

The last post completed the process part of the project.  I’ve tried to summarise the process on the attached diagram / graph.  A bit difficult to summarise: too many target audiences and interests to work for all.

On the diagram a timescale runs along the bottom x axis, starting at -9h to depict “Preparation” the day before.  The scale is not linear.  Up the left hand side of the graph is temperature as the y axis, focused on the 24C target but with an ambient temperature of around 29C and substantially cooler ingredients. 

During Preparation we can see the temperature of the ingredients is changed to get to the temperature we want at the start of mixing: the example here involves warming the flours and cooling the water. 

After Preparation, we move onto Mix: from time zero we have mixing to create the dough with the goal of reaching the target process temperature of 24C: the red dot.  Note the time scale change from hours to minutes.  This is followed by around a 3m period Scale when the mixed dough is formed into balls and tubbed.

Then the ball Maturation period follows, keeping the temperature constant at 24C.  In the example this takes us up to 11h.

At this point the dough has doubled, x2, at the start of the Window of Consumption.  The first ball is now used to make pizza.

The y axis only serves in this example for the Preparation and Mix aspects.  In the future if temperature was varied during Maturation, modifications would have to be made to the process.

Now ingredients...

Pizza Ingredients

Flour, water, yeast and salt first, then toppings. Even though some toppings may be unusual, alternatives with different names exist: perhaps this is more about flavour combinations, than specific names.  As part of the project ambition, I have attempted to use ingredients from Spain and Internationally.

Pizza Base: Flour

With a lot of exaggeration, we find 100% Caputo Pizzeria 00 flour produces a cake or pancake like character: very light, creamy, but too much “give” even though the protein level is high compared to a regular Type 00.  Alternatively, we find 100% Spanish T-55 strong flour (protein 13%) can give a bread like character: far too tough.  I'm struggling to find a consistent way to express this, the distinctions are very small.

Using Spanish T-45 and T-55 general purpose flour (8.5 - 9.5% protein) gives us a good texture, creamy, somewhere in between the above, but the dough is too weak to easily handle: especially as some of our topping recipes deliberately use overloading.  We expect to start using this flour more in future blends but it is not a priority.

When the first sack of Pizzeria arrived we made poor pizzas as the hydration was too low: 55% was ridiculous, but some recommend it.  The third session had 60% hydration and was amazing: it handled so easily, and a "creamy" flavour.  So we were very impressed, and carried on happily. 

A year later we tried a blind comparison session, half of the dough used Pizzeria, the other half a Spanish general purpose / German strong blend that had been popular the year before.  Hydration 62%, yeast and salt the same.  It was almost impossible to tell the difference.  The difference between pizzas made using the same flour was greater than the difference between the different flours.   What had changed however was there had been an intervening year of progressively improving technique.

We were left with the conclusion that Pizzeria was more tolerant of a less experienced or less patient pizza maker.  Our old "pre Pizzeria" blend however was a bit unforgiving of poor handling: it had a higher "big hole in the pizza" rate and some dodgy corniciones.

We continue to try different flour percentages, sometimes varying with mood and appetite.

The spring 2019 favourite was a blend of 10% Spanish strong flour 13% protein, 30-40% Spanish strong flour 12.2% protein, and 50 - 60% Pizzeria.  Then for the project it was 10% Spanish strong flour 13% protein and 90% Pizzeria.  We have moved away from Pizzeria since, now we are more experienced we can get close to what we want without the costs.  We believe successful marketing image to be a large part of Italian flour prices.  As a lot of Italian flour is made from imported grain, it would make great sense to find a way to differentiate product value and increase sale prices: the base ingredients being a commodity.

The attached table shows some of the successful home blends we have used to date: the table calculates the overall protein % for each blend, we currently stay above 12.2%.  There are several blends yet to try out ("n" in the used column), especially using more general purpose flour.  The colour coding indicates unusually low protein levels: we have had success down to 9.5% in the past. Most of these blends refer to pizzas made before the full standardisation of the process.  It’s a work in progress.  You may have spotted that the bottom right hand 12.2 figure is an error, it should be 12.28: apologies.

Blend protein content is just one aspect as the number hides many things such as the proportions of different proteins.  We have found flours can be a challenge once you go beyond the packet label.  At "European level" flour and additive health aspects are regulated, but each European country has its own specification standards: different to USA standards.  In local supermarkets flours from different European countries are available, we have seen flours from 4 different countries stocked on the same shelf length in a local supermarket.

Italian, German and Spanish flours are all specified differently, some have huge allowed variation (ie Italian 00), some have different analysis moisture contents so ash and protein figures are not comparable, some have very tight specification (Germany), some have unique measurement methods (Spain).  France and Germany do not specify flour types using protein, they use ash content.  And of course the French are different again. 

The country of wheat origin has to be specified in many countries, also the country of milling.  Country of wheat origin is not specified on Italian flours, but in Spain it is a selling point: grown and milled in Spain for instance, as much wheat is imported (Italy also being an importer of wheat).  Here we get into commercial blending, common in the USA, but many European flours are milled "straight": a single wheat ground and packaged as one.  Pizzeria is a blend however.  Anecdotally, German flours are much more finely ground than USA flours for the same protein level.  Whatever “same” means.

It gets worse: unlike the prairies of USA and Canada, the provenance of wheat strains is highly variable in Europe based on geography, climate, much smaller scale and consistency of farms, size of field, let alone variations within and between crop/s.  Quite independent of the choice of internationalised sowing seed stock and fertilisers. 

We don't worry about all of this any more.  Now we just use protein % as a loose indicator of character and stick to 12.2 – 12.5% blends. 

The Pizzeria is purchased from Napoli in 25kg bags, it is immediately broken down into 3kg ziplock food bags with as much air as possible removed.  The bags are stored in clingfilm sealed 30 litre plastic lidded tubs which contain around 100g of coarse salt to fix air humidity.  The tubs are kept on a tiled sealed concrete floor in the base of a cupboard in our basement.

Over a year there is a 19-25C temperature range in the basement, with a very slow rate of temperature change.  25kg of Pizzeria lasts for 11-15 months, and we have almost finished our third bag, with no degradation noticeable (checked by making a batch of remaining old flour and new arrival flour dough together, and blind tasting the pizzas).  The flour was purchased in July, August and September over almost 3 years, so the flour in each bag will have been milled and blended from grain at least 10 months older.  There is 3kg remaining of a Pizzeria sack bought in July 2018.  It's fine.  As you’ll see later, it may be there for some time.

The strong flours used in the blends are bought throughout the year in 1kg bags, usually buying several bags at one time from the same batch.  We use supermarkets with high turnover of stock.  The bags are stored in a floor level kitchen cupboard, simply because these flours are rapidly used with bread making: throughput is around 4 - 6kg (9 - 13 lbs) of flour per month.  These flours are sourced from Germany (wheat grown and milled) and Spain (wheat grown and milled) and are unbleached with no additives.  The brands have been used by us for some years.  How much variation is there ?  No idea, no indications to make me suspicious.  We have also used brands which do not work well.

Pizza Base: water, yeast and salt

The pizza water is bottled spring water (Sierra del Segura Neval Spring, Murcia) with 243mg/l sodium potassium and magnesium bicarbonates, 122mg/l sulphates, and 147mg/l chlorides.  It tastes very pleasant: definitely not too hard, but with a bit of edge.  There are some bottled waters here which taste a bit insipid, others too mineralised.  This one suits us and is really cheap in 8 litre bottles: it is our house general-use bottled water and is widely available in Spain.  We have not experimented with water hardness and solute levels:  if there is a good reason to do this this, please update me.  We know that water is a key variable in brewing.

The yeast was “Mercadona” supermarket brand IDY (no additives or processing chemicals) sourced from Germany.  The yeast manufacturer is unknown.  German purity laws are relevant, this is probably why there are no additives to the yeast.  On completion of the project we changed source to one of the big yeast suppliers: Lallemand.  I digress for a mo:

Lallemand products are not just Canadian: they are widely available in Europe through their acquisitions and manufactured in several European counties.  Mine is a 500g bag of Instaferm 01 (Red) IDY from the acquired Portugal factory.  Sorbitan Monostearate is also present in this yeast, as a residue from the manufacturing process.  It is also used in pile ointment. 

NOTE that the SC strain/s used in this IDY is/are unlikely to be the same as in the same product made in North America, as it may be tuned to Portuguese and Spanish bread manufacturer and market preferences.  Spain, France and Germany make significantly different mainstream and artisanal bread products with distinctive processes. 

Whilst not really part of the project, we have found that the new yeast has changed the dough Growth Rate upwards to around 0.13h-1.  ....an IDY yeast from a different manufacturer made using a different process giving different results....  to be expected.  The yeast is definitely from a different supplier / process as when Sorbitan Monostearate is present in a food in Europe it has to be declared by law.  I need more pizza sessions to refine the new Growth Rate estimate adequately to compare with the controlled sessions.  As you will find at the end of the next and last post, that is unlikely to ever happen. 

The salt is fine sea salt locally sourced, nearby are the salinas of some of the biggest producers of sea salt in Europe.  The salt is not iodised and does not have any free-flow additives.  Please note that so-called rock salt is actually sea salt too.  All that varies between one unadulterated salt and another is the proportion and type of contaminant.

We do not add anything else to our doughs.  We have a range of other more artisanal flour types at home, but have not experimented with these in pizza.  We do not like adding oil ( we have used EVOO up to 2%).   We avoid sugar for health reasons and may in future try reducing salt levels further, again for health reasons.

Margherita Time

The recipes presented in earlier posts give an impression of "excessive" toppings.  Our pizza sessions in reality are dominated by the basic standard pizzas, with some experimentation.  The perfect Margherita is clearly a personal experience: whilst we respect the perspective of the AVPN, we are global in outlook and do not consider their perspective to be a worldview.

Below are our variants on Margheritas which we call the Mutantes.  We enjoy “straight” Margheritas, but the twists below are popular with guests, and one in particular we believe is a genuine improvement.

Mutante Uno: Add Jamon Serrano: the usual scrunched up pieces, 6 -10 pcs.  The saltiness offsets the bland cheese perfectly: but that really comes down to how salty the tomato sauce is.  We don’t detect a difference when we use expensive jamon as opposed to the supermarket standard.  This raises a question of how salty a pizza sauce is: professional chefs tend to use a lot of salt as they know it “enhances” a dish.  Ingredients vary hugely in the amount of salt they contain: compare a Parmigiano with mozzarella for instance.  Different cultures tend to use more salt than others, and palette is personal and trainable.  I use as little salt as I can get away with.  One of our guests automatically smothers everything in salt.

Mutante Dos: No mozzarella, use Tierno blends of Spanish cheese with or without Emmental.  These are all low in salt compared with mainstream cheeses.  We have played with clumping or discrete lumps of cheese.  I prefer lumps, others prefer clumps.  Also the issue of whether it either melts out to a flat surface or not, forms grease droplets, or picks up char.  Another taste and visual appeal issue. (Pictures attached to make the point)

Mutante Tres: Use fine leaf basil, not broad leaf.  It’s commonly available here being more hardy in hot dry conditions and it continues growing outdoors in winter.  The attached picture in the pot is from the start of January outdoors, and the flowering stalks from October.  The result is quite different and every pizza bite has some basil in it with a very heady perfume.  Occasionally we have visually confused oregano for small leaf basil.  Attached is a comparator pizza picture with both types of basil.

Mutante Cuatro: No Pizzeria 00 flour, flour being one of the German / Spanish straight flours or a blend (see earlier on blends).  This is partly about texture: our eaters say cake lovers like 00, chewy sour bread lovers like T-55.  But I have control: I make the blends and my eaters are laboratory guinea pigs being tested every week.   Nothing in it.  Frankly, we have given up trying to tell the difference: there are easier tricks to play such as time and hydration. 

Mutante Cinco: Decent Spanish tomatoes.  Aha… a lead into Toppings….

But which of the Mutantes delivers the most ?  It’s personal to some degree, but there is a consensus on a winner.

Pizza Toppings

Below is topping information additional to that provided in the recipes.  Apologies for any duplication. 

We constantly experiment with ingredient substitution, not being receptive to the AVPN viewpoint when it gets to toppings.  The Mediterranean is a big place with wonderful ingredients throughout, just as every region in the world is, in the eyes and palette of the individual.

We vary the proportions of the cheeses per individual pizza to match the strength of the topping flavours.  The Emmental is a low cost “industrial” production general purpose cheese, not a high end version.  Cheap mozzarella prices start here at the price of Tierno mixture and Emmental. 

We have available locally a range of mozzarellas, we have tried more expensive Italian variants, including with buffalo milk.  We were surprised and a little disappointed.  Marketing ?  Cooked cheese on a pizza does not taste the same as when eaten uncooked at room temperature.  We really enjoy eating mozzarella uncooked.

The Gran Padano  is usually replaced with old hard Spanish cheeses.  (Parmigiano is a variant of Gran Padano, the latter more widely available in Spain)  The Spanish cheeses are usually blends of sheep goat and cow milk. 

We finely HAND CHOP all of the hard cheeses: a cheese grater provides pieces too thin for our topping finish:  we want to see melted tiny “blobs” of these cheeses on the finished pizza for taste, charred or not, and appearance.  Otherwise it is invisible and just provides umami.  The old cheeses vary noticeably in taste, and cost around 60% of the price of a cheap Gran Padano.  They are around 25% more expensive than Tierno.

If you look at the 2019 World Cheese competitive rankings (which the US won with a cow cheese), you will see a dominance of Spanish cheeses which are predominantly sheep and goat (three of which we use entirely by coincidence).

On the subject of umami, we have occasionally sprinkled, and I do dare say it, monosodium glutamate (MSG) onto tinned Peruvian artichokes and "bland" batches of onions during topping preparation.  Likewise we use celery salt rather than sea salt when preparing toppings, for that extra "je ne sais quoi".

Pepperoni, being a US product, is not available here: the Spanish are particularly proud of their own spiced and plain sausage types, chorizo, salchichon and loganiza being well known. 

Chorizo is widely used as a generic name outside Spain and Portugal, but even here there are many different varieties here which can be hugely different.  The name by itself gives no idea of what the sausage will be like, other if it is from Spain or Portugal there will be pork and pimenton in it.  A pork product.  I have heard of a Mexican chicken chorizo.

Sausages here are raw, air cured or smoked, air cured from the hot south, smoked from the cool north.  Often we find them too greasy to use on a pizza.  We use a very finely sliced moderately spicy chorizo with care, crumpling up individual pieces so they do not lie flat on the pizza, and can catch char on their edges.  We have a favourite loganiza which is high in pork content and low in fat, we pre-fry that to get maximum char.  We have not found salchichon or morcilla (black pudding) to work well. 

Jamon: enough said on that earlier.  We saw a TV programme where a single small curing attic had more than 8 million Euros worth of jamon curing.  Crazy stuff.

Vegetable toppings: the only ones which are not allowed are sweetcorn, carrots and parsnips, sweet potatoes and regular potatoes.  We did try potatoes with skins. 

We have a huge range of fresh vegetables locally with extended growing and cropping seasons.  Two of our supermarkets have local distribution centres so their fresh produce is only a few days old, the other chains have national distribution centres in the middle of the country, we do not buy from them as it may be up to 5 days old.  Local markets are open most days and produce is often a day old.  A big cooperative close by sells “seconds” from its packing centre as you watch the produce come in from the fields.

BUT once the vegetables have been toasted on the pizza, they are no longer fresh... and often look like distant relatives of the pre-cooked topping.  So does it matter if they are a few days older ?  I don’t know.

A general principle we follow is to DRY OUT toppings when they are being prepared, to reduce excess water.  This includes queso, cebolla, pimientos, alcochofas, calabacin, alcaparrones, guindillas, aceitunas and lacon.  Certain toppings we generally precook and season (salchica, embutido, bacon, langostinos) some toppings are prepared both precooked and raw (cebolla, pimientos, champinones).

The prepared toppings (and sauce) are removed from the refrigerator a few hours before use to WARM UP.   

A question - why put a home pizza with a wet cold topping into the oven ?  Illogical.  We are at home not in a restaurant.

We find that certain topping combinations fit different pizza disc types, cooking temperatures and duration: reflected in the recipes.  For instance, a spicy meat based pizza (caliente pizza) works best for us with a minimal cornicione, and a good degree of surface char.  A pizza with a raw egg added (fiorentino, desayuno pizzas) has to be cooked at a low temperature: especially as the entire topping except the egg has been pre-cooked.  The egg is usually just skinned over at the end, very runny.

Pizza sauces

We have two sauces, one is a tomato base, the other a mixture of the tomato base with one of sour cream, cream cheese or greek yoghurt, to make a rose sauce we call salsa rosado.  The proportions vary, we mix to taste: we seem to be putting in less and less tomato into the rosado over time.  The tomato base has evolved and has been pretty stable for the last 2 years. 

We have gone through the usual experience: local fresh pera tomatoes (pear tomato: what you may call a plum tomato) and cherry tomatoes, many different tinned varieties, and some cases of DOP tins from the San Marzano valley.  Also the preparation of the tomato salsa: straight out of the tin or fresh slices, hand crush, potato crusher, hand mill, stick blender, variable speed food processor blender….  Then the flavourings to put in the salsa.  So what did we learn about tomatoes ?  A personal view:

The Mediterranean countries are full of pera tomatoes and EVERY country, let alone region, claims they have the best tomatoes.  Some of the best tomatoes we have had come from the former Portuguese colony of Goa (now in western India).

The Italian tomato variants "of pizza tradition" were established in, and came from Spain: Spain and Portugal spread tomatoes all over the world. 

For instance, one of our local markets has entire individual stalls selling just tomato varieties, often with the same variety (usually pera) having differing prices depending on age, taste and the amount of soil on them (they are grown close to ground: plants are in the soil, not hydroponics).  The quality of fresh pera tomatoes in our region is so high, it’s called the salad bowl of Europe. 

I am absolutely sure the same is correctly said in southern California, perhaps even more so.  Good tomatoes are not exclusive: they are a global phenomenon (except for those Dutch hydroponic things).

In our opinion, fresh tomato slices or cherry tomato halves are sadly too wet to use as a topping.  The cherry ones also tend to fall off when placing the pizza in the oven (my problem).  Chunks of peeled, de-seeded fresh tomatoes work, look good, but are largely tasteless on a cooked pizza.  Preparation pain with no eating gain.

Spanish tinned whole peras need some “water” removal to make a sauce we like.  They vary substantially by brand.  Some of the cheapest brands of tinned pera tomatoes are far, far better than expensive ones: you just have to shop around to find a good batch !  I vaguely remember reading a comment like that in this forum some years ago, it was spot on.  We have a favourite brand.

Typically with a 800g tin (large) I strain off the liquid with a sieve, let the liquid stand for a while and remove 100 - 150g from the top of the liquid as it separates and clarifies.  Then recombine the rest and use it.  This liquid tastes GREAT though: used as a consomme in Michelin level cuisine… and by us.

The DOP San Marzano’s were poor: huge amount of juice, fleshless toms.  What is the fuss about ?  Clearly the AVPN is there to promote the interests of Napoli: as a former marketeer I thoroughly support that.  It may be that we did not have ‘the best’ San Marzano’s, but buying DOP: they have to deliver a level of quality.

We avoid tinned diced tomatoes: they are too coarse and need water removal then blending, but cost more.  Not the best sauce, for us a consistent failure from one brand to another.

We have settled on a Spanish tomato triturado, which is a tinned de-skinned de-seeded crushed tomato paste/sauce coarser than an Italian passata (passata means mashed, but the ones I have bought in several different countries have been very fine, almost a Bloody Mary mixer).  The triturado has not been dehydrated, it contains the usual 0.5% salt and inevitable citric acid.  It is stocked in all local supermarkets as a TOP seller, usually on palettes in 800g tins.  We have a favourite brand.  If you dilute it with around 30% water and blitz to make it finer, but not aerated, it makes an evil Blood Mary base too with a bit of added green olive jar juice.

To the tomato we add celery salt (essential, a devastating condiment) freshly ground coarse black pepper and sometimes, a little dried basil or oregano, EVOO and Jerez vinegar (aged sherry vinegar, similar to Balsamic for flavour without the added sweetness: a lighter subtle character).  Generally we run with just the salt, pepper and dried basil and put it in the blender for a minute on a VERY slow speed to make it a little, not much, finer so it spreads more easily.  In 2019 nearly every week it has been the celery salt and pepper only.  The sauce stands for several hours before use after the blender.  I am not convinced of the effect of adding anything other than celery salt and pepper.

We have found that no matter how we prepare it, a sauce made from fresh tomatoes (skinned and de-seeded) cannot remotely compare with a good tinned tomato.  Ah well.  Sometimes the truth is not what you want it to be.  At least the vitamin content is similar.  Tins are tops.


Offline Arne_Jervell

  • Registered User
  • Posts: 888
  • Location: Norway
  • I Love Pizza!
Re: Project Complete…2y
« Reply #43 on: January 05, 2020, 07:05:51 AM »
Stephie thanks so much for a dedicated, passionate and thorough writeup. I think your approach is fascinating, it is far from anything I've ever read before.

You have got things pretty much dialed in, yet I would not be surprised if the "complete" part of your project is a temporary state.  :-D

Enjoy your pizza and keep posting!  :chef:

Offline stefan

  • Registered User
  • Posts: 56
  • Location: Stuttgart
  • I Love Pizza!
Re: Project Complete…2y
« Reply #44 on: January 06, 2020, 06:09:37 AM »
Hi Stephie,

I really enjoy your work here! As Arne already mentioned, I think this is one of the most scientific approaches I ever saw regarding our topic! And, of course, also your pies look delicious! ;)

I have similar experiences regarding DOP San Marzanos - nearly all of them which I tried were quite bad. The only good ones I used were Mutti San Marzano, but as I don‘t get them nearby I mostly exchange them for regular Mutti or Cirio.


Offline stef

  • Registered User
  • Posts: 33
Re: Project Complete…2y
« Reply #45 on: January 06, 2020, 01:30:46 PM »

Thank you for your kind words on King’s Day. 

So here’s the final installment, but there is a Post Project Post on the way too.

Part sixteen of the Pizza Project:

Costs and concluding words

This is a good point to share costing information, straight after toppings.  I don’t have an unlimited budget, and am health conscious 5 years into retirement.  So I have not just done costs, but ingredient nutrition calculations.  Europe has mandated food product information labelling, I guess similar to what the FDA does in the USA.  For every ingredient I use, I have cost per 100g, food information per 100g, weight of product used per pizza, so for every ingredient and pizza combination I have a “health breakdown”, and the cost of the healthiness.   Crackers eh ? 

See attached picture for two flour labels, one a German Spelt (Espelta, Dinkel) flour, and the other a Portuguese Rye (Centeno) flour.  These are not wheat: everything looks comparable until you get to azucares, fibra and proteinas (sugar, fibre and protein).  I’m sure you get the idea.  Why did I really do this ?  Two of our guests said pizzas are unhealthy.  Sure, if you do a Papa John a day.  In the interests of potentially your health, I have completely left the healthy stuff out.  But pizzas aren’t bad.  I also do the RDA calcs (Recommended Daily Allowance) for the killers, sugar and salt.  Onto Costs...

Fixed costs

Capital one off investment (hopefully) in equipment: oven, mixers, fermenter, oven tools, measurement tools.  These have been expensive but the likely life expectancy is very long.  The oven depreciation could be over 10y, the tools 5y and the mixers 12m.  For some, the oven could be considered as adding to the sale value of the house: or not.  Cheap measurement tools are accurate, but perhaps need to be handled with care: almost disposables in the overall scheme of things.  Mixers are very expensive.  But we all know the world of enthusiast pizzamaking is littered with the detritus of broken mixers. 

My total oven cost is less than the Bosch and Kenwood mixers together.  Clearly for some, oven costs can be exorbitant, ours was low because of the substantial self-build element.

Variable costs

Fuel: we are seeing payback on our decision to have a moderate size oven with very heavy insulation (insulation was the second biggest part of the oven total cost): our average pizza session uses up to 15kg of wood (33 lbs), burnt over a 2 - 3h period.  Wood is low cost here as it is a primary home heating fuel.  Wear and tear costs to the station wagon when collecting wood are probably higher than the wood cost itself.  So wood is a low cost element of our total pizza price.

Flour: The Italian Pizzeria 00 costs 2 - 3x that of German and Spanish flours (that much variation over three separate bag purchases).  To the time of writing we have used 24 different flour blends, some of those blends have been used more than 20 times.  Excepting pizzas made with >10% very strong flour (13% protein, that is v v high for Europe) and more than 70% general purpose flour (9.5% protein) the Growth Rate is constant and it is very difficult to impossible for my eaters to tell the difference.  I can tell the difference with >10% very strong (poor taste and dough handling, significantly higher Growth Rate), also >50% general purpose (very delicate taste but difficult dough handling). 

The Pizzeria is “on the way out of the door”, it has served its purpose superbly and we thank it for that.  Nevertheless, we will see that even that is a low cost element in the overall pizza price. 

Toppings: in the estimates both the expensive flour and cheap flours are used to proportionally define the range of cost differences.  Expensive flour = Italian, Cheap flour = German and Spanish.  This is done so RELATIVE costs can be seen: no point in giving Euro prices if you are in Australia etc.

Below a range of indicative relative costs are given for topping ingredients compared to an average flour cost per pizza.

Topping ingredients were costed by ingredient item / pack price (very variable item weight for vegetables, some for cheese, none for tomatoes), and scaled to 1kg price to allow comparison between items, and dividing it by the weight typically used per pizza.  This gives a the cost PER PIZZA of every topping we have used.  For seasonal price variations I have done rough averages or nothing at all. 

Costing the toppings has been the most surprising aspect: in particular cheese, fish and meat. 

The basic cheese topping (Tierno, Emmental, Mozzarella) is 2 - 4x the flour cost per pizza, the hard cheese sprinkle is 0.5 - 1.5x the flour cost.  You’ve got it: a sprinkle of cheese can cost more than the base.

Longostinos, prawns, ham, sausages are expensive: between 4 - 10x flour cost. 

Vegetables are between 0.2 - 2x flour cost, preserved vegetables insignificant as they are cheap and get used over a long period for a variety of uses, tomatoes averaging 1x flour cost, occasionally up to 3x.

Log costs run at around 2 - 4x flour costs per pizza ( ie 2x in the case of expensive flour, 4x in the case of cheap flour).

Overall an average total topping for a pizza is around 13x the flour cost, lower and upper limits of between 8 and 17x flour cost.  This is a very rough estimate.  This excludes any depreciation of the capital costs.

The bottom line is that topping cost makes fuel cost and flour cost look low in comparison.  What is more important: an exotic set of toppings or a fabulous base with utilitarian toppings ?  We are not a restaurante.  Let’s remember the Margherita and Marinara are low cost pizzas.

Concluding words

It's useful to "start the end" with a reminder of the Pizza Project ambition:

    • Simple and predictable process
    • Understanding cause and effect
    • Data collection for improvements
    • Work with the climate: south east Spain
    • To use ingredients from Spain and Internationally
    • Use precise measurement: especially temperature and weight
    • Reduce costs: I am retired

Looking back over the posts, these have all been covered to some degree.  So concluding observations are:

I enjoy pizzas more than ever.  Measuring dough does not decrease pizza appetite.  I have dreamt of pizza.  Many times.  Conclusion is: the process is somewhat addictive...

Keeping a Log Book with notes and detail of every pizza session proved invaluable: in writing the posts I have frequently gone back to specifics from July 2016 onwards.  I won't bore you with the spreadsheets though.  Conclusion is: if you record everything, one day the records will make the pizza for you.  (If you ever read them.)

At the start of the project, "I did a Cortes".  Whilst having got away with it here at least, the downside is that the people who were told about the project are now QUITE sure that "I am different".  I’m just surprised that they didn’t think this before.  Conclusion is: pizza changes other people’s perceptions.

This amateur science and process based view of domestic pizza, removing guesswork, caused much, much more work than anticipated.  I did not know whether or not the above ambition could be completely achieved at the start of the project.  Conclusion is: domestic pizza can be predictable, and still very tasty.  Even more tasty.

I have a baseline from which I can now make improvements, ONE variation at a time.  Those of you who have read this far probably know what I mean by a baseline and have spotted a few of the tricks.  And one change at a time recorded so you can roll back.  Conclusion is: I am starting to sound like a systems model.

Growth Rate and Lag Rates work.  Enough on that one.

End to end temperature control works.  It is straightforward once you have set up with some "dry runs", it then makes everything more predictable and reproducible.

The short within-day fermentation times made my tolerances tighter than may have appeared necessary.  The brevity of the "Window of Consumption" really focused the mind upon getting it right.  The tight window has really driven the project, else it would not have addressed core points such as Lag Phase and dough temperature. 

Lessons learnt with the short process seem to be partly applicable to longer fermentations, but that is not for now.

Simplification took place throughout the process: whilst some detail has been presented here, this has been a "one off" project feature as part of getting a greater understanding.  After all, a project is defined as a "one off" activity.  The detail is a one off and is now filed.

The detail has little bearing on the pizza process today: I don't recalibrate other than the odd spot earring check, I don't go into detail on yeast Growth Rate or FDT.  I just use a couple of equations in my tablet spreadsheet.  The recipes generally change on the basis of what is available on the day.  There is one message here though: you cannot simplify unless you know what you are simplifying: detail counts.

On pizza costs, the biggest have not been where expected as flour and fuel costs are very low here, even premium flour brands in comparison with toppings.  Toppings have  been more expensive than expected: the cost reduction opportunity is there, but there also lies a lot of enjoyment.  On capital costs, that has been expensive but it's a long life and depreciation term.  My conclusion is that really good toppings are worth paying for, the other costs are "down in the noise": a surprise.

Nevertheless.... there remain elephants in the room: natural and commercial ingredient variability.  Whilst that was out of the control of the project, it’s an inevitable conclusion: the effect of that variability is unknown and not domestically manageable.  Reality.

A last word on variability and accuracy.  Variability is often an enemy of science and engineering, fought with accuracy.  Hopefully a reader of this thread may know a little better why pizza making results vary so much.

Finally , the arrival and handling of the dough after it comes out of the tub onto the counter, then oven management.  Deliberately, this has not been discussed.  Why ?  Because I am delighted to live with the joys and the pleasure of making and eating the pizza itself.  See the attached recent white hole.

Oh and nobody spotted the deliberate hole in the FDT discussion: the elephant in the mixing room was carefully hidden by a deliberate project decision: to hold hydration constant, even though some doughs come out stickier than others.  Check out how your FDT varies between 70% and 55% hydration. 

That's all for now:  apologies for being boring and confusing.  I did want to share, and hope that some of the words may be of some use to you. 

Most Important, I gratefully acknowledge all of the expertise of the individuals who have shared on this forum.  Thank you so much.  You really wrote this.

Now the girl who ate pizza stays awake at night not thinking about fermentation times, but other things instead.

Final recipe time.

A word on the star of the pizza: artichokes, or alcachofas: we live close to the “heart” of the Spanish artichoke crop: it's in season as this is written.  They have been grown in the Mediterranean for thousands of years, naturally as cardoons then the cultivar significantly developed in Andalusian Spain by the Moors, where the Arabic and Spanish name Alcachofa originated.  Small green artichokes are so popular here you even see cans and jars sourced from Peru.  We can't grow enough of them.  I do not use fresh artichokes on the pizzas, I use tinned or frozen then cooked.  With the tinned ones I let them air dry once quartered for as long as possible for the preserving liquor to evaporate.  If you squeeze them out to dry you lose some of the flavour.  Tearing leaves off fresh artichokes both discolours my hands for a week and breaks my nails from time to time.  Fresh artichokes are wonderful… but not for pizzas.  This recipe is about the artichokes.


Casa Farsha Pizza Dos

Base of tomato: the regular base, 70g
Tierno cheese: irregular small clumps, cut into the usual pieces, 45g
Artichokes: 12 pc place regularly, drained & cut into quarters, 30g
Lacon: 8pc placed regularly, cut to small size (2cm x 2cm), 25g
Black olives: 8 pc place regularly, brine drained, pitted and halved
OPTION: 8 pc jamon serrano if you need bigger flavour: this is intended as a subtle pizza.
OPTION: 8 pc immature goats cheese if you need bigger flavour.
OPTION: Courgette: sprinkle all over, matchstick cut, salted, water removed, 15g
“Oveja Viejo” hard mature sheep cheese: irregular sprinkle, medium / fine chop, 5g
EVOO spiral
Big twist of black pepper.

PS there is a Post Project Post on the way….

Offline Icelandr

  • Supporting Member
  • *
  • Posts: 1490
  • Age: 73
  • Location: Gabriola Island, British Columbia
  • Infinite Monkey Theorem at work
Re: Project Complete…2y
« Reply #46 on: January 07, 2020, 11:16:15 PM »
Thank you Stef! As I have said earlier, my mind doesn’t work that way, it is a very good thing there are different approaches and skills. I read most, perhaps glossing over more of the scientific fact gathering than you would like, but have seen a side of pizzamaking I hadn’t really considered, you lost me at cost and calories for each 100 grams of ingredients and relating the cost offsets for your vehicle when gathering wood, but you won me over and I performed a little Dance Of Joy when you confessed to dreaming of Pizza . . . . Now we are talking the same language!

Very well done, a bucketload of work for you and yours, carefully researched and well presented, Thank you from the other side, All the best!
PizzaParty 70x70, saputo floor

Offline stef

  • Registered User
  • Posts: 33
Re: Project Complete…2y
« Reply #47 on: January 09, 2020, 01:26:31 PM »
This is the Post Project Post.

It's an update to where we are now 3 months after starting to write up the Project.  It also gives an indication of where we may go in the future months (other than a very warm location further south).

This will be brief as it’s only being re- written once: the Project text had been re-written 5 times, hence all the qualifications you will find if you read it carefully.

New World of dough

For the last 3 months we have completely abandoned using any Pizzeria 00 flour, and have replaced it with 100% German Type 810.  This has been used with the Instaferm IDY yeast. 

Our pizza results have been great, good strong dough with exceptional lightness.  If you look at it the dough too long, it may run away.... being difficult to handle at long fermentation times: but we are down at 12.2% proteins.  From a Fermentation Time perspective, the Growth Rate has changed

Attached are New growth curves at 24C, with an extended time axis up to 50h.

The uppermost and lowermost curves are the Golden Chalice and Craig TX curves as used previously (green and purple respectively). 

The blue and red lines are my curves from Growth Rate and regression curve fitting in the within-day Project sessions, which were looked at in detail in the posts.

In previous posts these curves featured a lot, also discussion of how to reduce the differences between IDY% and Fermentation Time curves. 

The black line and crosses are my last 13 pizza doughs in the recent 3 months.

To summarise:

My black curve now sits half way between Chalice and Craig TX: all that has changed is the yeast manufacturer and the flour blend.  In itself the flour change does NOT make noticeable difference, its been used many many times before. 

These results are why I “subtly threw out my handbag” about the issue of yeast CONVERSION RATES, and the industrial design of the yeast itself: it was actually spotted during the project write up.

The power regression curve fit shown on the graph is crazily high: R2 = 0.99759.  With more data points that surely will reduce.  But I suspect the rigour of my process has tightened up through practice.

HOWEVER, the bottom line with all this stuff about yeast and fermentation is that the SHAPE of the curves remain the same and DO NOT converge. 

So a tip for an interested reader: find a few good data points from your pizza records and you will be able to extrapolate
up and down with confidence by following the shape of the curve. 

So my model going forward is:

%IDY = 0.79581 * (Fermentation Time) -1.09607

Finally, more than 18 months ago I regularly used 100% Type 810 at 62%.  I NEVER got results as good as with the recent pizzas.  Technique ?  Process ?  As was pointed out by Craig TX: more experience has to be the big part. 

New World of toppings

Since the summer of 2019, the Tierno goat, sheep and cow mixture has been replaced by a 100% goat Tierno cheese.  This is a little more expensive then the mixture but seriously delicious, and the melt and char characteristics are the same as the other cheeses, including mozzarellas.  It is verging on bland, but has panache.  It is not salty in comparison with a Feta.  Part 15 has a pizza picture with this cheese: can you tell which picture ?

Even better, it’s a local goat’s cheese from Jumilla, Murcia (better known for Jumilla wine, which featured in the Arroyo-Lopez et al (2009) good paper on SC growth rates: but for vino not pan). 

We have also found a local Lorca, Murcia goat cheese producer who regularly wins World Cheese Gold and Silver Awards with the most amazing super tangy umami mature cheeses, to replace the other Parmigiano substitutes. 

Essentially, we have gone native.  It was always part of the ambition, and for us it wildly exceeds expectations.  My only message to a reader is try goat cheese: young and old.

This led to the next step: use our home grown lemons and olives: pizza photo attached: a 34h dough with home preserved lemon and home cured olives, home grown basil with local red peppers, spinach and goats cheese; but the jamon serrano from 300km away.  Too dry here for pigs.  The cut red pepper was previously microwaved for 4m to soften it and get some char: a “dirty” trick to shortcut putting it over a flame to get the skin charred. Beware you can easily burn it in the microwave…

Our olives are tiny, there were around a dozen on the pizza.  Tasted gorgeous though, the preserved lemon was surprisingly in balance with everything.

The broad leaf basil is out of favour.  It looks great, but it’s not ideal to grow here with the dry heat and winter’s cool nights.  The small leaf basil (fino) is tough, and we know there will always be some to pick at any time.  It also seems caterpillar resistant, unlike the broad leaf which has to stay under cover to keep the moths off at night.  Or it might be that ants eat the moth larvae: the fine basil is always full of ants, which give the bonus of providing a certain crunchiness. 

We have tried using the fino both traditionally on top of the cheese and heretically before and below the cheese.  The end result surprisingly looks much the same.

I prefer the taste and perfume but not the look of the fino.  Try out small leaf basil.

New World of Temperature Control

The new doughs were fermented in my new fermenter.  It’s still being modified and tested (an early test picture attached, since a water bath has been installed)  That has given me the confidence to work to longer times at 24C, I have so far gone up to 62h in tests, but the real plan is to study the effect of changing the temperature of a fermenting dough: especially the rate of temperature change.

However… whilst my dough balls are coming out at x2 at the right time, the dough has become EXTREMELY difficult to use at these extended times: it’s crazy dough.

The dough does not need forming into a disk, it almost runs off the granito by itself.  A slurry, like a collapsed over-fermented dough.  I have not been able to tell the difference between a 34 and 48h dough: at all.  I’ve made the pizzas, they have been GREAT to eat.  No fun to make. 

Clearly, something is breaking down: I’m not an organic chemist and am sure that others can instantly explain what is happening given the 24C temperature.

On the other hand, 48h at 24C gives approximately the same growth as 140h at 17C, 30h would be 90h and 12h would be 36h at 17C.  Obviously the comparison is not ideal as different reactions will be occurring, and at different rates.  Using tubs, my 62% x2 needs careful handling compared with those Youtube videos in any case.

Looking at the graph, the doughs at approximately 26 hours were fabulous: yes they try a little to escape off the granito but they are worth the care needed to make the disk.  Using a bench flour mix of 40% rice and 60% general purpose flour stops too much flour being needed, the “put in the fridge an hour before disking” trick too… and really it’s clearly just more practice. 

My Chief Taster decided these pizzas and the 21h ones were a MAJOR improvement over the within-day doughs.  Light.  Tasty.  FYI, that was NOT their reaction with the 24h doughs we used to do at 17C, which were definitely bland and undistinguished, and with almost monochrome leoparding (both Pizzeria and Type 810).

Going forward, 20 hours is a walk in the park.  The upper limit looks like being up to 30h.  I will find out.  There will be a lot more of these longer doughs.  I will also add 10% of the 13% Spanish flour to hold it together better, bumping up the protein to 12.3% ish.

Attached is a "stack" picture of why there will be a lot more: note the colour of the base is beige, not white, and the lower slices are being squashed by the weight of the upper ones.  Club sandwich, eat your heart out.  Another slice picture shows not just the beige colour in comparison with the white cheese, but a crispy brown outer crust.

So the journey continues.

New Project: from Concept to Plan

Previously I had decided that the complexities introduced by changing the temperature of dough were too much of a challenge for a newbie to take on.  But I know a little more now.

So the girl who ate pizza but no longer stays awake at night thinking about fermentation times... is now staying awake refining a Concept which will lead to a new Project Plan.

Attached is the end to end Dough Process diagram used previously, as a reminder.  Also attached is the diagram adapted to make my Level 1 Concept for the new project.

The new diagram has compressed the early stages (Preparation, Mix, Scale) and final stage (Consumption), and extended the timescale along the bottom from <13h up to <60h.

We are still assuming an ambient temperature of 29C, a FDT of 24C and only 200g dough balls to minimise dough bulk temperature issues.
The dough maturation will change from the 24C used to date: to both 21C, and 18C (both 3C drops).  There will be a Ball Maturation at both of those temperatures preceded by a Cooling period (1-6h) and succeeded by a Warming Period (1-6h).  So there will 2 sets of runs, at 21C and at 18C, with varying cooling and warming periods. The 21C and 18C  temperatures are shown on the vertical axis in blue and red.

The fun starts looking at the lower part of the diagram with the dotted lines, all under the category of Growth Behaviour.  This is what I’m after.  Running through each in turn:

Lag Phase: expected to be around 10 – 12h based on data from earlier posts and factoring the temperature reduction.

Lag T change +20%:  my shorthand for the additional Lag Phase duration caused by the instability due to temperature reduction during Cooling.  This is expected to be around 20% of the Lag Time at the target 21 and 18C temperatures: possibly around 2h.

Growth Phase: The Total Fermentation time minus Lag Phase.

CO2 to Solution: Once Growth Phase has started the CO2 generated during yeast growth that remains in solution rather than forming bubbles, because of the reduced temperature increasing the Partial Pressure of CO2.  If the dough was warm this would gas out and create actual growth, but this dough is now significantly colder.  The effect of this is to delay the onset of actual dough growth, even though the yeast is multiplying in Growth Phase.  Duration unknown, but possibly calculable.

Actual growth at maturation T Rate: Once the dough liquid is saturated with CO2, this is the perceptible growth through to Warming time: the Rate of Growth will be that expected for the temperature, probably around 0.08 - 0.1h-1  (NOT the end to end rate).

CO2 to Gas: Once the dough enters the Warming time and starts to warm, the CO2 in solution (accumulated earlier) will gas out as the Partial Pressure reduces.  This has the effect of boosting growth beyond that expected for the Growth Rate corresponding to the dough temperature.  This is what accentuates leoparding, the extra gas bubbles against thin dough walls.  Leoparding does not occur if the dough is held at a constant temperature.

High Growth (T Rate + CO2): this is the net Growth Rate during warming:  T Rate is shorthand (sorry) for the increase in Fermentation Rate as the dough increases in temperature, from say from 0.08 - 0.1h-1 to 0.16 to 0.19h-1 . This is accelerated by the CO2 to gas release occurring at the same time.
I am not using bulk dough for reasons discussed at length in earlier posts: too complicated.

The Cooling and Warming periods will be varied using the new fermenter.  Temperature control will use the Inkbird and 2 sets of external pumps to a water bath in the fermenter, one providing hot and the other cold forced liquid circulation coolant, the balls being in immersed stainless steel containers: this is to achieve dough cooling and heating rates between 17minutes /C to 50minutes /C.  (the 1-6h cooling and warming periods) The bath will also have its own continuous circulation pump.
The fermenter is a modern high quality refrigerator carcass encased in 8 – 20cm of styrofoam, contained by shaded 6cm solid walls and concrete floor and ceiling.  At various stages of testing I expect an additional 12 thermocouples to be in place too.

But we are still at Concept Phase, with a lot of ground work to do.  I expect to drop the 18C runs as there may be too much work to do. If it stays I may drop the FDT temperature.  I am also sure some aspects of measurement will prove too difficult.  But one change at a time.  After more prototyping I will then set the Ambition and Design.
I will maintain as much as possible in terms of rules from the original project, it is likely the German 12.2% flour will be bought ahead as a single batch with possibly 20% of the dough blend being 13% Spanish, the water, yeast sugar and measuring tools will be held as is.  I am considering using the Bosch mixer instead of the Kenwood as I perceive a more consistent end of mix dough character.

Ah well.  the dangers of being retired...

Thank you all again so much for your help and support  !


Offline amolapizza

  • Lifetime Member
  • *
  • Posts: 2013
  • Location: Luxembourg / Spain
  • If pizza is food for the gods, what are we..
Re: Project Complete…2y
« Reply #48 on: January 09, 2020, 02:24:18 PM »
The new doughs were fermented in my new fermenter.  It’s still being modified and tested (an early test picture attached, since a water bath has been installed)  That has given me the confidence to work to longer times at 24C, I have so far gone up to 62h in tests, but the real plan is to study the effect of changing the temperature of a fermenting dough: especially the rate of temperature change.

However… whilst my dough balls are coming out at x2 at the right time, the dough has become EXTREMELY difficult to use at these extended times: it’s crazy dough.

The dough does not need forming into a disk, it almost runs off the granito by itself.  A slurry, like a collapsed over-fermented dough.  I have not been able to tell the difference between a 34 and 48h dough: at all.  I’ve made the pizzas, they have been GREAT to eat.  No fun to make. 

Clearly, something is breaking down: I’m not an organic chemist and am sure that others can instantly explain what is happening given the 24C temperature.

IMO, it's the gluten web breaking down.  It's too long at too high a temperature for the strength of the flour.  Lower the temperature, shorten the maturation time or increase the strength of the flour and maybe increase the salt level.  Maybe lowering hydration level also helps.

One might consider why ferment/mature the dough to the point where the gluten is breaking down.. On the other hand even if the balls are hard to handle, they make a tasty and interesting pizza if one can manage to cook them without disaster :D

I'd like to thank you for this series of articles, very interesting to read!

I spend a lot of time in Spain, and think that with the quality of the raw material available there is little need to use imported ingredients.  Next time I make pizza in Spain, I'll try queso tierno.  I often eat it fried for breakfast :)

Effeuno P134H (500C), Biscotto Fornace Saputo, Sunmix Sun6, Caputo Pizzeria, Caputo Saccorosso, Mutti Pelati.

Offline amolapizza

  • Lifetime Member
  • *
  • Posts: 2013
  • Location: Luxembourg / Spain
  • If pizza is food for the gods, what are we..
Re: Project Complete…2y
« Reply #49 on: January 09, 2020, 02:32:50 PM »
This link might be helpful! https://harina.info/

Effeuno P134H (500C), Biscotto Fornace Saputo, Sunmix Sun6, Caputo Pizzeria, Caputo Saccorosso, Mutti Pelati.


Offline foreplease

  • Lifetime Member
  • *
  • Posts: 7496
  • Age: 61
  • Location: St. Joseph, MI
Re: Project Complete…2y
« Reply #50 on: January 18, 2020, 09:51:49 AM »
Thnk you for this most interesting topic - both your work and dedication to it. I’m only through post 18 but looking forward to continuing.