Avoiding oxidation of Extra Virgin Olive Oil: A lot more than just dark glass
admin | April 8, 2010It’s often said that the olive oil industry has strong parallels to the wine industry. Well this may be true in some areas, but in the case of managing dissolved oxygen levels to maximise the shelf life of the product, the respective industries are worlds apart.
Most winemakers know the importance of minimizing dissolved oxygen in their product, particularly at the bottling stage. It is well known that bottling white wine that contains anything over 1mg/L of dissolved oxygen (that’s 1 thousandths of a gram per litre) will lead to a wine with an economically disastrous shelf life. But for some reason many extra virgin olive oil producers don’t seem to be overly concerned about dissolved oxygen levels in their product. Sure, most know that you should strive to always have full tanks, and that you shouldn’t pump around unless you really have to – however, oxygen pickup by EVOO occurs at every stage of the process – from crushing to bottling, and depending on what packaging is used, even post bottling.
Extra virgin olive oil is blessed with natural properties that afford it limited protection against oxidation: Sure, it’s dominant fatty acid, oleic acid, is relatively resistant to oxygen attack, and it contains lots of natural antioxidants including polyphenols and tocopherol. However like all defenses, constant attack by oxygen will eventually wear the oil down, tiring it and finally killing it with rancidity.
One thing is indisputable though: If the olive oil doesn’t contain any dissolved oxygen then it can’t oxidise. Ok, completely protecting oil from oxygen pickup isn’t a feasible option – EVOO is produced in a world drenched in oxygen. However, research has given us some insight into just how much oxygen can dissolve into EVOO, at what stage it gets in, how much typically gets in, and what effect it can have on its long term stability and shelf life.
But first some general facts…..
Olive oil is very receptive to oxygen. One litre of olive oil can absorb up to 35 milligrams of oxygen at room temperature. In contrast, at the same temperature, one litre of white wine will become saturated with only 7 to 8 mg of oxygen.
When fully saturated with air, the 35 milligrams of oxygen contained in the oil is sufficient to increase its peroxide value of the oil by up to 2.5 units. This is a significant amount given that most good quality EVOO’s have peroxide values around 6.
The oil will quickly consume any oxygen dissolved in it and then readily take up further volumes of oxygen. The process will be repeated as many times as the producer will let it. Oxygen consumption by the oil only leads to one thing in the long run – the dreaded food destroying rancidity.
What about processing?
Processing of olives into oil will naturally result in oxygen being absorbed and dissolved..
The production of extra virgin olive oil using the modern continuous process naturally results in a dissolved oxygen content of around 8mg/kg which may increase peroxide levels by around 4 units.
The relative contributions of the processing steps: malaxation : horizontal centrifugation : vertical centrifugation in terms of oxygen pickup is approximately 1 : 2 : 3.
So despite popular belief, the majority of oxygen pickup does not occur during malaxation. Blanketing malaxers with inert gas in fact has a relatively small effect on oxygen pickup. On a related note, some oxygen pickup during malaxation is necessary if the EVOO is to develop maximum flavour. While fully blanketing malaxing paste may help increase polyphenols levels that doesn’t necessarily relate to EVOO quality.
And Tank Filling?
Filling olive oil into tank without any oxidative protection results in 80-85% oxygen saturation, which results in increased peroxide values of about 2. Yes, saturating EVOO with oxygen is that simple!
Sparging the oil with 99% nitrogen (that is bubbling it exhaustively through the oil) at the recommended rate will reduce dissolved oxygen levels 10 fold to around 2.5mg/kg (with a resultant decrease in potential peroxide value from 2 units to around 0.15 units).
Sparging does not afford complete protection as the inert gas used to sparge usually contains some oxygen. It is also worth noting that sparging has the potential to remove some of oil’s aroma and flavour volatiles.
And Bottling?
Consider filling a 500ml bottle with olive oil without the use of inert gas. The 500mls of air that the bottle contains, comprises 142mg of oxygen – that’s enough oxygen to saturate the oil 4 times over!
Adding inert gas to the headspace of a bottle of oil after it is filled has a relatively small impact on dissolved oxygen levels. If filled to around 99% of its total volume, a typical bottle will have a 6ml headspace which contains 1.8 mg of oxygen. This oxygen has the potential to increase the peroxide value of the oil by a maximum of 0.2 units. (as using inert gas only partially displaces oxygen from the headspace due to mixing effects, and that inert gases are not entirely pure and can contain oxygen themselves). Every bit helps, but if the oil is already saturated with oxygen at bottling, headspace topping with inert gas is probably more of a feel good exercise.
Oxygen pickup can also occur when bulk oil is stored in plastic bladders in the mill, or when bottled for retail sale in plastic containers. This is a very complex area due to the variety of different storage materials that are either currently available or are under commercial development. Something to leave for another day.
Hi Richard,
In this and several other posts, you mention the importance of dark glass bottles to protect EVOO from degradation by UV. I had wondered if tin containers were not even better, as providing a complete block of light, but had recently begun to wonder if (on the analogy to people getting a significant dose of copper or iron from using cooking utensils composed of these materials) the oil might pick up some transition metals, leading to *more* peroxidation and polyphenol loss due to Fenton reactions.
By coincidence, I just had an experience that certainly gives *one* unexpectedly-dramatic, albeit indirect, demonstration of the potential for this to happen. I had bought an EVOO in a 1 L tin online, and was certainly pleased by its fruitiness and complex flavor. But after finishing it I removed the top with a can opener to recycle the tin and saw that there is rust on the inside, going down about a half-inch from the top lip of the can. (If you’d like, and have a mechanism on the blog or would take an email attachment, I can send you a photo).
My *guess* would be that the demarcation line indicates where the olive oil had reached when the can was full, and that the rust occurred because of residual moisture in teh top of the can. Now, at the very least, this means that there is a significant amount of oxygen at the top of *all* tins of this brand (although not much, and you seem to suggest in this post that this is only a minor factor in oil degradation under normal circumstances). But also, actual *rusting* will clearly mean that iron oxide particles are getting into the oil, presumably sinking down to the bottom (and then mixed around every time I shake it up …), accelerating such problems dramatically.
Do you have any idea how common this phenomenon is? Is there any way to be sure that it *isn’t* happening? The oxygen could be avoided by either flushing with nitrogen, or by ensuring an even more complete filling (which would also drive out more moisture), but it also seems that the integrity of the tins themselves isn’t great. And, as I say, even before this event I had been concerned about transition metals leaching from the tin, even absent rust. I don’t normally see this in tins of food; is there something one could ask a manufacturer about the tins themselves (which are normally *called* “tin cans” because they have tin plating inside the steel, in part to protect against corrosion!) that would give one confidence that neither routine leaching, nor actual rusting, were occurring inside the canister? Perhaps your lack of mention of tins on the blog is intentional; should one just not buy EVOO in tins, but only dark glass?
Thanks again for your informative blog!
Hi Michael
Thank you for your very thoughtful and extensive comment. While I don’t pretend to be an expert on tin packaging I do agree with your assessment. Photo-oxidation is the most destructive of the oxidative pathways. And despite popular belief, dark glass does transmit some UV radiation while steel cans do not. However you are right, iron is an effective catalyst for oxidation so if the tin lining is compromised in any way then you have a nice pathway opening for oxidation.
Steel can manufacturers have in recent times made advances in reducing the weight by reducing the guage. This is a positive thing as it reduces the energy required to produce and recycle the can and the amount of energy required to transport the can from producer to consumer. However, common sense suggests that lighter guage cans will be easier to compromise.
I was suprised that the can had corroded above the high tide mark. While olive oil contains a small amount of water as does the air where the tin was filled. But to rust? Makes me think that the oil had been in there a long time or the can was in some ways defective.
I’ve been researching olive oil packaging and will write a blog on it. Hopefully shortly.
Richard G.
Hi again Richard,
Thanks for your followup on packaging in the form of different tints of glass. Assuming that you’re still interested in doing a more comprehensive review on packaging, you might find this book chapter on Packaging and the Shelf Life of Vegetable Oils (from the book fFood Packaging and Shelf Life: A Practical Guide ) useful; also likely its principal reference on EVOO, How the choice of container affects olive oil quality – a review . (In the process of finding these, I came across the Aussie study Effect of Storage Containers on Olive Oil Quality, which wasn’t very useful for our purposes an also didn’t show up your countrymen terribly well …).
Looking forward to more of your informed commentary.
Hi Michael
I’ve been thinking of writing something on packaging but have been frustrated with the lack of any decent controls in the studies I’ve seen. For example you come across comparisons of olive oil stored in a large tin container with a much smaller clear PET container whereby the oil was kept in the light. Well derrr of course the oils stored in cans end up better. I only wish that researchers in the field made a 500ml dark glass bottle with a tin ROTE lid the control in their studies and kept the oils in a dark cool place (you know the conditions which we all say EVOO should be stored in).
I’d also like to see good oils being used. How many times have you seen a study where the base chemistry of the oil is an FFA of 0.6% and a peroxide of 18, or something like that. i’ve always felt that if the researchers used a half decent oil then they would get more relevent results.
Re the Aussie study. I’m not sure how tanks or bladders being filled to only 10% capacity and lids left off and stored in the light at ambient temperatures relates to commercial practice. But I’m sure the researchers had their reasons. In any case, the study was about storing oils in 1000L containers in the mill rather than what happens in your or my kitchen.
RG
Hi Richard,
Is it safe to assume that a cork stopper, however artisinal-looking, is a bad idea, as it will permit oxygen diffusion and thus increase the peroxidation rate in the sealed bottles?
Thanks!
Hi Michael
Cork is a natural product and as such is very variable in the amount of oxygen it lets in. Wine researchers have found that the range of oxygen transmission by good quality corks (called reference 2’s) was between 0.0002 and 0.1227 ml of oxygen per day (thats a 600x) difference. Screw caps allowed between 0.0002 and 0.0008 ml per day (a factor of 4). Given that the closure is simply that, something to stop the oil getting out, then I think that the screw cap stopper should be the preferred option, simply because of its consistent performance.
RG
Hi Richard,
I’m guessing you’ll find this paper useful, tho’ it’d’ve been nice if they’d tested at ‘fridge temperatures and with more measures of the same variable with all packaging materials.
Re the Aussie study: I agree it’s of little obvious relevance to home storage, but it’s alarming about what seems to be some fairly widespread practices in the Aussie EVOO industry:
Some samples of olive oil submitted to the Australian Oils Research Laboratory (AORL) for testing as extra virgin olive oil were found to contain very high levels of peroxide. Further investigation found the oil had been stored in plastic vessels, in the sun for several weeks. Information obtained from members of the industry indicated that these storage vessels were commonly used for storage and transport of high quality olive oil as well as for other types of edible oil. The manufacturers of these containers provided the AORL with a range of product specifications, showing that some containers were recommended for oil storage and some were not. Despite this, it appears that oil producers will often use the lowest cost containers despite the risk of product quality deterioration.
It also occurred to me that the show oils you’ve reviewed for comparison to IOOC standards might be fresh-bottled, vs. this stuff left for bottling later in the season.
Thanks for the cork info, as for other insights.
Dear Michael
Introductions in research papers are renowned for over-dramaticising situations to make the research seem indespensible and money well spent. That lab gets thousands of samples each year from hundreds of producers. So when they say ‘some samples’. were stored in the sun for several weeks.. well my guess they found one or two. ok one or two too many, but in any industry there are always those that don’t have a clue.
I was at the last Australian olive expo where this paper was presented and my feeling from the questions and comments from the audience that most felt that the treatments applied were in the extreme end of what may happen rather than what does actually happen. There were a lot of comments along the lines of ‘who actually empties out 90% of the oil from the tank, leaves the lids off and then puts them out in the sun’. Fair enough questions.
My personal view is that the treatments applied did in no way reflect standard industry practice, but instead extremes in bad practice. I’d also be surprised if plastic lined containers aren’t widely used for transport and storage by small producers in other countries as well. I have seen them used elsewhere outide of Australia while on my travels, but couldn’t say how much they are used.
All I can say about our show oils is that they are bottled and labelled oils which are available for retail sale. So their chemistry reflects what the consumer is getting. . Can’t do better than that.
RG
Hi Richard,
Following up on the detailed breakdown of likely contributors to oxidation: you wrote:
The production of extra virgin olive oil using the modern continuous process naturally results in a dissolved oxygen content of around 8mg/kg which may increase peroxide levels by around 4 units.
From this, and the rest of your post, I take it that even if one is using the best available tech, it is essentially impossible to get an oil that starts with a PV significantly under 5? Or can you drive it down with some other tech but there’s a tradeoff with some other production value?
You also wrote:
Adding inert gas to the headspace of a bottle of oil after it is filled has a relatively small impact on dissolved oxygen levels. If filled to around 99% of its total volume, a typical bottle will have a 6ml headspace which contains 1.8 mg of oxygen. This oxygen has the potential to increase the peroxide value of the oil by a maximum of 0.2 units. (as using inert gas only partially displaces oxygen from the headspace due to mixing effects, and that inert gases are not entirely pure and can contain oxygen themselves). Every bit helps, but if the oil is already saturated with oxygen at bottling, headspace topping with inert gas is probably more of a feel good exercise.
Now, subsequent to posting the above came Stefanoudaki et al, 2010, which I know you’ve seen because it was you that alerted me thereto 😉 :
In this study we have examined the effect of olive oil storage outdoors [“for 4 months in winter “] on a comprehensive series of quality measures. The conditions used were at the extreme of those encountered during the production of bottle oil. … Increases in K232, K270 and peroxides over time were very much reduced by inert headspace gas, which also reduced losses of total phenols and oxidative stability. Headspace nitrogen also reduced the rise in unconjugated phenolics as secoiridoid derivatives declined and minimised losses in polyunsaturated fatty acids. Panel tests were used. All oils lost perceived quality on storage and this was accelerated outdoors while headspace nitrogen slowed the deterioration significantly.
Now this is about the long-term oxidative stability rather than the initial PV which gets the ball of self-propagating peroxidation rolling, and having not looked at the full text I don’t know about the effect of light (was this “outdoor storage” in daylight??), or whether they found (as they seem to be saying) that they also saw these phenomena in indoor-stored oils (presumably at room temperture). But they are reporting a significant protective effect, whose magnitude I would obviously like to know. Based on this study, would you say that nitrogen flushing the bottle headspace would be a useful practice to improve shelf stability under normal (indoor, room temperature, and preferably (but in stores, not typically) light-shielded) conditions?
Thanks as ever!
Hi Michael
From what I’ve read, the decanting process is a significant source of dissolved oxygen. I’ve been told that some of the newer models of continuous processing machines have addressed this (to what degree I couldn’t say), and a few innovative processors have made engineering ajustments to existing machines that help as well. So, yes I’m sure you could produce and oil with a PV less than 5. Rod Mailer and his group in Wagga have reported seeing an oil with a PV of 1… but interestingly of the 155 EVOO’s submitted to this years Perth Royal Show, none were less than 5, with 5 being the minimum. Sparging oil with Nitrogen will drive off dissolved oxygen but typically also drives off aroma volatiles.
Re you second question. The devil is in the detail. The researchers used cans, so no light was involved, but the cans did heat up to 38C at times – as you would expect if you left tins of olive oil at ambient outside temperatures.
More importantly, they filled the 500ml cans to only to 460ml , leaving an 8% headspace, which would explain the bigger increases in PV than my theoretical estimate based on a more typical 1% headspace.
RG
It seems that using olive oil for frying is a subject few agree on, some say it will go toxic others say it won’t. I got to admit I’m lost on this on and will stick to the advice chefs give and tha’ts to put turmeric in the oil when cooking to stop it going rancid.
Hi Mike
Its pretty simple really. If you heat olive oil (like any other oil), it will go rancid faster than if you didn’t. In a domestic situation I’m not sure why anyone would want to reheat olive oil or any other oil multiple times.
The golden rule is JUST DON’T BURN IT!, as that is when you start getting nasties formed. But what I don’t get is why olive oil is singled out. The same applies to all oils.
I haven’t heard about the tumeric thing, but it wouldn’t surprise me as it has lots of powerful antioxidants in it. You would have to wonder about the taste that it would impart though.
RG