What I thought of the proposed Australian/NZ olive oil standards
admin | March 6, 2011I’ve been asked by a couple of people to post my comments on the proposed Australian/New Zealand standards for olive oil.
I’ve previously posted my thoughts on the general thrust of what they were trying to achieve here;
http://www.aromadictionary.com/EVOO_blog/?p=786
But here are the specific comments that I provided to the committee for consideration:
Chemical Standards
Many of the chemical standards are disappointing.
FFA should be reduced to 0.5%. This is double the Australian average of EVOO’s submitted to Australian olive oil shows (from micro-volume to >10,000 litres) and would adequately cover well made oils.
Peroxide value should be reduced. EVOO’s with PV’s of 13 or more are generally rejected by oil buyers as their shelf life is known to be severely reduced. Why should consumers be expected to accept the oils that others reject?
The DAG limits are set way too low. While the data set presented by Mailer et al. was small it would appear that even some of the worse oils in the marketplace would meet the proposed 30% limit.
Alkyl esters are not well tested and should not be included in the standards. An extensive search of the literature in academic databases show very few peer reviewed articles on the method and the number of samples used to demonstrate the validity of the method are severely limited in number and in scope.
The DAG and PPP methods are both useful chemical indices. The claim by some that the methods are ‘dynamic’ is difficult to reconcile given that the methods are used to assess a dynamic property of oils – aging and heating during storage. Furthermore, most of the commonly applied chemical criteria for quality such as peroxide, UV measures and to a lesser extent FFA are also dynamic in that they change with oil age. The other claim that they give a number of false positives is not supported by any published data.
I support the higher levels proposed for campesterol. It is based on solid published evidence on the natural variations in Australian olive oils and elsewhere. High campesterol levels have been shown to exist in Cornicabra from Spain, Barnea from Australia and Koroneiki from Greece and Australia. In fact Rivera del Alamo et al. (2003) concluded that “a large proportion (between 25 and 85%) of the annual production of this commercial monovarietal virgin olive oil exceeds the upper limit (for campesterol) established by the current EU legislation”. This level of false positives resulting from a 4% limit is unacceptable. So if false positives are an issue in the minds of the IOC then they should review their level for campesterol with a view to increasing it. Details of the samples (number, varieties, agronomic factors etc) that were used to set the current 4% EU limit (to my knowledge) have never been made public for scrutiny.
4.11 Free fatty acidity Minor point but “expressed as percent of free oleic acid” has been shown to be confusing even with even industry insiders claiming that high oleic acid levels are a negative aspect of olive oil. Suggest “expressed as oleic acid equivalents”. This is standard terminology when a complex mixture of something (in this case free fatty acids) are given as if they were a single compound (in this case oleic acid).
4.20 Peroxide value. Technically peroxide value is not a measure of the oxidation of the oil in that oils that are in a more advanced state of oxidation have already used up their peroxides and therefore can have a lower peroxide value than an oil that is not yet oxidised. PV is more of a measure of the oxidative precursors in an oil and therefore reflect its propensity to oxidise.
4.22 Pyropheophytin A Should read “… from thermal or age related degradation”
4.24 Sterols Not all sterols are “characteristic indicators of authenticity”, so this clause should read “ many minor constituents of oils, some of which are are characteristic indicators of the authenticity”.
4.26 Triglycerides should read Triacylglycerides (TAG’s)
Sensory Aspects
Clause 4.15.1 – 4.15.5: Could be simplified and combined into the following. I feel that there is no need to state where the defect comes from in these standards as many of the causes are sometimes speculative and/or not fully understood.
The current wording is too restrictive. For example, musty can be mushroom like but it is also often described and mouldy hay/dusty/old books etc. There is also an inconsistency in the proposed standards where some defects are partly described in terms of their sensory characteristics, while others are only given via their possible causes.
Furthermore there a variety of defects which whilst rarely seen can cause serious reductions in oil quality. For example the earthy character most commonly assigned to European oils as being picked from the ground is in Australia most likely the result of dirty wash water resulting from constant re-use in situations where the environment is dusty. Dried woody flavours are most likely the result of desiccation due to drought conditions etc.
I have also experienced a number of off flavours that are not given in the classic European listing of defects whos causes are unknown but may be the result of modern packaging and storage methods. These include “band-aid/medicinal” which while being tentatively assigned by some to be a muddy character, may also be the result of contamination with cresol based cleaning products, and a variety of plastic taints which could be caused by storage in inappropriate containers. Aerosols have also been reported to impart a solvent like ether character to the oils via their propellant. Therefore a more general wording to the clause is required so that it covers a more comprehensive set of possibilities.
Therefore I suggest that Defects be consolidated into:
“Any flavour that derives from enzymatic degradation, fermentation, or microbial spoilage of olives prior to processing, fermentation of olive matter following extraction, subsequent excessive oxidation, or any other character that could not be reasonably be assigned to the natural flavours derived from the olive. These include, but are not limited to the defects known as fusty, musty, winey, rancid and muddy sediment.”
6.1(a) The median level of defects equal to 0 means in (more simple terms) that the oil is ‘free of defects as determined by the majority of trained tasters’
Labelling Requirements
12.3.2.1 The limitation on suggesting that an oil has special characteristics where those characteristics are common to most oils could in the long term be detrimental to the olive oil industry. Other edible oils will still be able to make these statements on their labels, so it could build a perception that olive oils may actually be inferior with regard to salt, cholesterol etc. I am intrigued as to what is the harm in making such statements.
12.3.2.2 To fill a potential loophole, “on the same line” could be changed to “immediately preceding”
12.3.7.1 from date of bottling should read “from the date of packaging”
12.3.7.2 This best by date clause is very important. However it does not provide sufficient (in fact any) guidance to producers as to how to assess oil durability given oil stability measure and fatty acid profile. I would also suggest that
a) FFA should also be included in the list of considerations as free fatty acids are more prone to oxidation.
b) Peroxide value should also be considered as a factor (for obvious reasons).
The standard as currently written is ill-defined and as such cannot be applied by producers nor policed by regulators. As such it has little worth. Whilst I appreciate the complexity of factors affecting oil stability, the committee should propose a simple method of determining a best by period (say a rancimat value multiplied by a factor to give months + some time added to accommodate different packaging types or sizes. This would have the added advantage of encouraging producers to use appropriate packaging for their oils.
12.3.9 The clause mentioned within does not exist. The note has been extensively covered elsewhere and could be deleted. The exception regards stating a single chemical indice. I believe that this policy is retrograde. The standard should encourage producers to give the FFA on their labels. Whilst not perfect it is a good indicator of quality and does not change much after packaging. It is almost certain that oils low in FFA will also have other good chemistries as FFA suggests sound fruit, rapid processing and hygienic practices.
Overall I am concerned at the lack of differentiation between what are essential aspects of labeling regarding quality and those which merely serve to differentiate oils in the marketplace i.e marketing constructs. Specifically the standard calls for producers to be able to justify statements made regarding the style and claimed sensory characteristics of their oils. These things relate to style not quality and therefore should not be included in the standard. Moreover, there is currently no mechanism for the sensorial profiling of extra virgin olive oils. The IOC accredited panels are only trained to rate the intensity of defined defects and also bitterness/pungency. While the latter may be equated to style ie. Mild, medium and robust, there is no consensus on how the median rating given by a trained panel can be related to a word like robust or mild. That is, on the 10 point scale where does mild end and medium begin etc. I also feel that being compelled to have a trained panel describe oils moves too far into the realms of marketing and product differentiation. It is also very impractical. Even if panels were trained to describe the complex fruit characters of EVOO there would not be enough sensory capacity to profile the requisite number of oils. Again as a back label description of an oil is merely a guide to the consumer (and most people take them with a grain of salt anyway), I feel there is no need to compel producers to justify their descriptions of their oils.
Grades of Natural Olive OIls
The use of three grades (rather than the 4 used in the EU) should be applauded. I cannot see the relevance of the ‘ordinary virgin’ grade in an Australian/New Zealand context. An oil would never be marketed as ‘ordinary virgin’ being refined instead, so in effect rolling it into a lampante grade makes a lot of sense.
6.1 It is unclear if lampante olive oils fit in the “natural” category as it states that “Natural olive oils fit for consumption without further processing include Lampante olive oils, however the definition states that Lampante olive oils are not fit for human consumption without further processing. This is a major contradiction under the current wording.
6.2(a) Refined olive oil. Should read “… including deacidification and deodorization”
The 0.3% FFA limit for this category seems very high. If the refining is done effectively the FFA should be close to zero as refining by definition should remove the FFA via the deacidification step. As the FFA of oils does not appreciably increase with storage, this value is somewhat mystifying.
6.2(b) Refined olive oil blend. The standard does not specify the maximum level of natural oil that can comprise the blend. So presumably an oil in this category could have a FFA much higher than 0.6% (i.e. 90% low end virgin + 10% refined oil would have an acidity > 2%). The high median rating for sensory defects given seems to accommodate this, but the FFA does not.
However in practice it is unlikely that any oil in this category would have any more than a small proportion of virgin oil. In fact I believe that there is a limit to the amount of virgin oil that can be blended into refined oils in the EU. This is logical as it doesn’t make sense to produce a refined oil blend that comprises a high proportion of poor quality defective virgin oil. If the point of producing refined oils is to present a product which is bland but does not display any defective characters then the proportion of refined oils in refined oil blends should be very high, and virgin oils low.
Therefore there should be a maximum limit to the proportion of virgin oil that comprises a refined oil blend.
If this is adopted then the median level of defect would need to be lowered. In fact refined oil blends should be defect free.
The points also equally apply to 6.3(b) and (c)
Other
The committee has made a serious omission with regard to labeling of refined oil blends (which include pomace oil blends). The proportions of virgin to refined oil should be a requirement on refined oil blends. Currently in the marketplace there are EU ‘Light’ oil labels that state “made from a blend of virgin and refined olive oils” Given that the proportion of virgin olive oil in that blend would be under 5% as per EU regulations, the phrase used on the label has the potential to seriously mislead consumers. The proportions of each type of oil in the blend should be given and stated in order relative to the proportions i.e. contains a blend of refined (97%) and virgin (3%) olive oils.
Packaging
In general I support the thrust to include packaging into the standards. As olive oil is a perishable product then it is only as good as the package that contains it. However I feel that the standards do not adequately accommodate possible future advances in packaging technology (i.e PET with UV blockers) nor alternatives to bottles and cans.
The note should read “impervious to light at wavelengths likely to cause deterioration”. Some materials may not be as dark as others but they may be better at blocking energetic light at wavelengths in the near ultraviolet and at 418nm, (a wavelength at which the highly oxidative singlet oxygen is generated from chlorophyll).
12.2 Container filling tolerances. Bag in box, bladder, tetrapak etc are not covered by this standard. They should be.
I hadn’t read most of the Draft, being focused almost entirely on the ‘headline’ chemistry, whereas you’ve gone through this in intense detail — thanks very much for taking the time to go through it in detail and give them (and you) a thorough analysis.
I note that both the Draft and you endorse the use of rancimat/oxidative stability index (OSI)/ for establishing shelf life. OSI is, however, based on conditions far removed from those to which oil is exposed in normal storage, and alters the chemical process substantially, making it highly prone to artifacts of the test itself and therefore can’t in my view be considered reliable; eg, Frankel (2010), Chemistry of Extra Virgin Olive Oil:
Thiswould certainly be more time-consuming and expensive than rancimat, but it seems clear that OSI is simply misleading.
Hi Michael
Sure the rancimat test has its limitations and critics. And I don’t disagree with them one bit. Like a lot of tests regarding EVOO quality it is a very blunt instrument. But no one seems to be able to come up with a viable alternative to it. If you think of it as a rough gauge of relative stability rather than some absolute measure then it has some merit. The point I made to the standards committee was that they hadn’t given any detail HOW the various measures/chemical indices known to influence stability could be used IN PRACTICE to determine shelf life. Activation energies, induction periods and “Arrhenius plots of the log of the induction period versus the reciprocal of absolute temperature” don’t mean a lot to anyone (including me!!) 🙂
The more I think about it, the more I believe that things like rancimat can only be used to derive a ‘relative shelf life’ rather than an absolute one. Lets face it, the shelf life of the oil in the bottle isn’t only dependent on the oil’s chemistry. It also depends on the amount of dissolved oxygen (DO) at bottling, the type of packaging etc. These are the things that the rancimat test doesn’t consider, as the oil is poured out of the bottle into the rancimat machine and the oil is heated and air bubble through. It doesn’t know where it came from! But I think that some research could be done that could specify an oil that under some defined conditions of DO, bottle volume, light transmission of packaging and storage conditions, has a shelf life of the maximum defined by the standard. What almost everyone has forgotten is that the shelf life can only be determined by some form of human assessment ie. sensory criteria like the maximum percentage of consumers who found the oil unacceptable at a particular point in time after being stored under the defined conditions. There is no other way of defining it as old oil is not microbiologically unstable so it all gets down to how rancid it tastes. That is, any chemical indice of shelf life must necessarily be related to perceived shelf life.
If you follow this line of thought, cranking down the temperature on the rancimat machine doesn’t address any of these issues. There needs to be a rethink of this matter, but who will do it?
I hope I made some sense here. Shelf life of EVOO is such a complex topic, I often confuse myself.
R.