‘The potential of RPR (reactive phosphate rock) for New Zealand’
I have been involved with RPR research and promotion in New Zealand for 40 years. I spent 17 years as a soil fertility research scientist with the Agricultural Research Division of the Ministry of Agriculture – what became AgResearch – including 3 years as Chief Scientist (Soil Fertility) at Ruakura. I designed and coordinated the ‘National Series’ of RPR trials, which ran over a period of up to 8 years on 19 farms throughout New Zealand in the 1980s. The trial sites were all deliberately selected to have below-optimum soil Olsen P levels, to more clearly show any differences in performance. Despite this, differences in pasture production between RPR and super were minimal – 0-3% on average – and had totally disappeared by Year 3 on low and medium P-retention (ASC) soils and by Year 5 on the highest P soils. The only exception was a low-rainfall (700mm), non-irrigated site at Winchmore in mid-Canterbury which
had also been over-limed to a soil pH of 6.4.
The simple solution to any initial ‘lag-phase’ with RPR was proven to be to use a blend of RPR and high-analysis P fertilisers such as TSP, DAP or MAP (and added S as required) for the first few years. Note that mixtures of RPR and superphosphate were shown to be not as effective for this purpose. Subsequent research demonstrated that where Olsen P levels were at or above optimum – as is the case on about 98% of dairy farms and about 75% of hill country farms in New Zealand – there were no measurable differences in pasture production right from the start of using RPR, meaning a a soluble P component is unnecessary.
If the fertiliser industry in NZ was not dominated by the superphosphate manufacturing duopoly, I am convinced that RPR – mixed with sulphur and other nutrients as required for individual farms – would (i) already be the main source of P used in NZ, and (ii) we would have far less polluted waterways and lakes as a result. RPR is proven to result in far less run-off of water-soluble P than superphosphate. Imported high-analysis would be blended in where required.
Also, because RPR particles dissolve in the soil steadily over time, releasing P for direct uptake by plants, there is far less P existing as soluble P adsorbed onto the surface of soil particles than is the case with superphosphate. Most of what is described as ‘particulate P’ lost from soil in run-off and erosion is actually present in the form of
soluble fertiliser P that has become adsorbed (to use the technical term) onto soil particles near the soil surface. When these particles end up in a waterway or lake through soil erosion, as much as 25% of this adsorbed P can easily be desorbed back into water-soluble form, (as demonstrated in an excellent soil chemistry paper by Australians Barrow & Shaw in 1975). This ‘particulate P’ form of loss is also greatly reduced with RPR, but you need multi-year constant-treatment trials to clearly demonstrate this.
Unfortunately, this vital area of water-quality research has received no funding in NZ, largely because most of the government research funding on fertiliser and the environment is – completely inappropriately – channelled through the duopoly, who have no wish to encourage this research. New Zealand simply does not need to be taking the manufacturing-grade phosphate rock from the Moroccan-occupied deposit in the Boucraa area of the Western Sahara to make into superphosphate. Certainly, superphosphate has played a very important part in developing NZ’s low-P soils. However, virtually all our agricultural soils have long been developed to the point where they can now be maintained very easily with slow-release form of P, containing up to 30% soluble P where needed, to ensure that our
waterways are protected.
If we continue to allow the industry to be dominated by two management groups who refuse to accept what is happening to our environment, we will only have ourselves to blame as we progressively lose our hard-earned reputation as a ‘clean and green’ country. Farmers must consider the question ‘Which P fertiliser should I use, and why?’ far more seriously than they have in the past.
Algerian RPR is easily the match of any other RPR agronomically, and contains a low cadmium level of 18ppm, which represents only 140mg Cd per kg P, well under the Biogro’s and Demeter’s organic farming limits, and only half the limit that the industry allows itself. It performed even better than North Carolina RPR in trials run by the
International Fertilizer Development Center, Alabama, USA. The Managing Director and the Senior Scientist of the IFDC released the following statement in 1999: “Unground Djebel Onk (Algerian) phosphate rock is classified as a
highly reactive phosphate rock for direct application to acid soils”.
All NZ soils are acid. As it happens, all RPRs are also liming agents in their own right, automatically reducing and in some cases going close to eliminating the need for maintenance lime applications. Algerian RPR has the highest lime equivalent (58%) of all RPRs, helped a bit by the small amount of naturally-occuring dolomite running through the
deposit (3-7% by weight). Note that the naturally-occuring 3-7% phosphatic dolomite in Algerian RPR can reduce its citric acid solubility in NZ’s current but obsolete 30-min test. This is an artefact only, and has no effect whatsoever on the excellent field performance of Algerian RPR. It also contains among the lowest levels of Cd, mercury (Hg) and uranium (U) of all IFDC-recognised RPRs.
Because of all these positive attributes, some industry players have tried to put farmers off using Quinfert Algerian RPR by playing the ‘it is not soluble enough in NZ’s test’ game. So we also offer the product with some of the dolomite screened out, to ensure it reaches 30% citric solubility in the current test. Both the normal (V1) and ‘low-Mg’ (V2) versions perform the same as each other in the field as fertilisers, i.e., exceptionally well. One just has a bit more dolomite than the other!
Finally, there are several other low-Cd RPRs available for blending from around the world as well, so there is absolutely no reason for anyone to resort to reducing the high cadmium level in Sechura RPR by mixing it with a low-Cd manufacturing -grade (non-RPR) phosphate rock, which may be as little as 20% as effective as an RPR.
A 50/50 blend of Sechura with Boucraa slimes (PB3) or Moroccan rock may be only 60% as effective agronomically as 100% RPR. And take note, the Khouribga rock from Morocco, commonly used in the manufacture of fertilisers, can contain large amounts of uranium, up to 566 ppm (FAO, 2004). This is so high (10 times higher than Algerian RPR for example), it can be economically mined to produce uranium! Ask for an updated declaration of the heavy metals in your superphosphate.
The ball is in your court. Please phone or email me if you have any questions.
Dr Bert Quin
021 427 572
Quinfacts 9. Quinfert Algerian now available in two grades.
Solely due to games being played by the industry, Quin Environmentals (NZ) Ltd has decided to make 2 very slightly different grades available to farmers.
V1 or ‘HRPR’ was field-tested by the IFDC, and rated by them as a Highly Reactive Phosphate Rock. It performed even better than North Carolina RPR in published response trials conducted by the IFDC. It contains 12.7% P, 35% Ca, 0.8 % Mg (as free dolomite present in the deposit) and 1.3% S.
Note however that the small amount of free dolomite present naturally in V1 (5-7%) slightly reduces its solubility in the obsolete 30-min citric P test (citsol test) used in NZ (and now nowhere else in the world, to my knowledge). This small amount of dolomite has NO EFFECT on its excellent field performance. Ironically, the very same people who mix Sechura RPR with Moroccan rock and call the mix an RPR are arguing that V.1 should not be called an RPR. Go figure.
V2. Is Algerian RPR from the very same deposit, with very slightly more P (12.75% total P) due to some of the dolomite being removed (magnesium content reduced t0 under 5%). This product meets the obsolete NZ citric solubility test.
If you wish, you can choose whether you want V1 or V2. Same price. They perform the same as one another. You get 0.2% more Mg in V.I or 0.05% more P in V.2. Your choice. Let us know. It’s all a bit trivial, don’t you think? I think so. But then, I’ve only been researching RPRs for 41 years…
Quinfacts 8: The time has come for New Zealand to completely stop making superphosphate
18 September 2018
The National Series of RPR vs superphosphate trials, ran for 3 to 7 years in the 1980s on 19 sites throughout New Zealand. The sites were deliberately chosen to have below-maintenance levels of soil Olsen P. This was done to make it easier to assess if RPR performed better or worse than superphosphate. There were tiny differences (typically 1-3%) on some sites in the first 1-3 years.
Where soil P levels are at or above maintenance (as is the case on 98% of dairy farms), no differences occur. Where soil pH is 5.6 or below, as occurs on over 80% of hill country, no differences occurred, regardless of the Olsen P.
The only situation where RPR had not fully caught up with super by year 4 was a non-irrigated, over-limed (pH 6.4) site with a very low Olsen P in dryland Canterbury (average annual rainfall 750 mm).
In any situation where a small difference in production may occur initially (called the ‘lag-effect’) there is a very simple, proven solution. This is, to use a mixture of RPR and a soluble form of P such as triple super, DAP or MAP, in a ratio that gives 30% of the total P in quick-release form, for the first few years.
So OK, it’s proven that RPR-based alternatives are just as effective. But why should we change from using super? There are a number of important reasons. I will just deal with some of the most important here.
Perhaps the very most important is the fact that fertiliser P losses in surface run-off after rainfall are far greater with super; there is virtually none with RPR. This P run-off loss is the prime cause of eutrophication of waterways and lakes. Some people who should know better have said things like ‘there’s no point; you still get what is called particulate P losses regardless of what fertiliser you are using’. But particulate P is largely soluble P that has become ‘adsorbed’ on soil particles. Using RPR long term, production can be maintained with much lower levels of ‘adsorbed’ P. The RPR stays present as RPR particles that slowly get dissolved by soil acid and used by the plant. These particles are far, far denser than superphosphate and therefore much less prone to being lost in run-off. Short-term trials on land that has a history of superphosphate use cannot be expected to show this effect. But not a cent of the tens of millions of taxpayers money that goes to the superphosphate industry for ‘research’ via so-called ‘Public Good’ funding gets spent on this. No prizes for guessing why.
Then there is leaching. Up until 15 years ago, most NZ researchers dismissed the idea of P being leached through soils. We now know that enormous quantities of P applied as superphosphate (up to 40% of it) can be leached right through the very low P-retention soils of Northland and the West Coast, and that significant P leaching occurs on all soils with P retention values below 45%; in other words about half of New Zealand. RPR by contrast does not leach.
And then there is the fixed amount of sulphur (S), present as sulphate-S, in superphosphate. Super normally contains about 9% P and 12% S if correctly made (ie fully acidulated). This an S:P ratio of 1.3 to one, almost double the 0.7 to one S:P ratio that pasture actually needs. The excess sulphate-S gets leached from the soil, taking with it (for electrical charge reasons) a mixture of the cations calcium, magnesium, potassium and sodium. These losses all have to be replaced over time to maintain production, representing on an annual basis a totally unnecessary cost of about $30/ha annually. Under high rainfall, almost all the sulphate is leached (taking even more cations with it), so the farmer has to use super that is fortified with elemental S, a form of S which doesn’t leach. All of this expensive, completely avoidable nonsense could be avoided by adding the precise amount of fine elemental S to RPR for each farm. No wastage, less cost, less leaching of cations.
And what about storage and manufacturing costs? RPR has a much higher P content than super (12.7% P for Algerian RPR) compared to 9% P (probably a maximum 8% P available). Combined with the much higher bulk density of RPR (1.65 compared to 1.1 for super), that means a given amount of P (or P plus S) as RPR or RPR/S requires only 40-45% as much storage size as does super. Only 65% as many tonnes need to be transported and spread, greatly reducing costs to the farmer.
Then there is the liming value of RPR. Every tonne of Algerian RPR applied has the liming action of 580 kg/ha of pure lime. This is offset by the elemental S added, which produces acidity as it is converted to sulphate-S by soil bacteria, but RPR/S containing the agronomically correct S:P ratio of 0.7 to 1 still has a liming value of 300 kg/ha. Therefore RPR/S being applied at say 275 kg/ha is providing, for free, about 85 kg/ha of pure lime effect. Most soils in NZ need 150-250 kg/ha lime expressed on an annual basis, but about half of this is a function of sulphate leaching from super. Put simply, farms using RPR/S will only need half the amount of lime to maintain a given soil pH than where super is being used. And the full liming effect of RPR of gypsum (calcium sulphate) is used as the source of sulphur instead of elemental S. The important thing is to only use the amount of sulphur that is needed.
Superphosphate did a good job helping to develop the productivity of New Zealand’s soils, but put quite simply, it is way, way past its use-by date.
Quinfacts 7: The facts about serpentine superphosphate
17 September 2018
- Serpentine is a natural magnesium-calcium silicate. It is widely used as a slow release magnesium (Mg) fertiliser in finely ground form. In the case of the product called ‘serp-super’, about 25% by weight of finely ground serpentine is added to the freshly-made superphosphate, before the sulphuric acid has all been utilised in reacting with the phosphate rock to produce soluble phosphate. This means that some of the Mg in the serpentine (maybe 20%) is converted to water soluble form. This may be a slight advantage in the short term, if soil Mg levels are particularly low.
- The big disadvantage of doing this however, is that the more acid utilised attacking the serpentine rock, the more manufacturing-grade phosphate rock, which is essentially useless agronomically, is remains in the final product. We saw in Quinfacts 6 how even in ordinary super, up to 15% of the total P can be present as unreacted manufacturing rock. In serp-super this can be 25%.
- So a serp-super advertised as having say 6.8% P, 8% S and 5% Mg, may in fact have only 5% usable P. No mention of this is made in information provided by the superphosphate manufacturers. In my view, this must constitute misleading advertising. The manufacturers cannot have it both ways. They cannot criticise reactive phosphate rock (RPR), which typically dissolves in the soil up to ten times faster than does a manufacturing phosphate rock like Boucraa (from the disputed area adjoining Morocco), yet very clearly imply that all the manufacturing-grade phosphate still present in the super you buy, and even more in serp-super, is plant-available. But this is exactly what they do.
- Finally, the question of asbestos. Asbestos is actually one of the many crystalline forms of serpentine that can exist in a deposit. Some serpentine in some serpentine deposits is present as asbestos; some fortunately contain none. When a serpentine deposit containing asbestos is mined and crushed, much of the deadly-to-lungs fine asbestos can be released. Much serpentine used to make serp-super in the past was made with asbestos-containing serpentine. More of the dust got released during spreading. It is now illegal to mine serpentine containing more than trace amounts of asbestos.
Quinfacts 6: Important facts about superphosphate and RPR
Superphosphate is advertised as having 9.0% P. But how much of this 9% is actually plant available? An indication of this is how much of the total P is citric soluble. Currently, it is about 8% P. The difference is largely unreacted phosphate rock that did not get acidulated in the manufacturing process. Because superphosphate is geneally made from hard manufacturing rocks, rather than reactive phosphate rock (RPR) , the value of this difference is questionable.
Another indication that unreacted phosphate rock present is present in superphsphate is to look at what level of sulphate-sulphur is present. If the phosphate rock is fully acidulated (ie, converted to plant-available form), the sulphate-S content should be about 12%. Figures of 10.5 or 11% indicate unreacted manufacturing rock is present.
The actual plant-available P level in superphosphate has a big influence on the cost per kg of P, especially on a transported and spread basis because it is a relatively low P-content product in the first place. At say $310 per tonne, and say $120 per tonne transport and spreading, if it really has only 8% plant-usable P, this means the true cost per kg P is about $5.37 per kg P. So it is important to know exactly what is in it.
By comparison, all the P in RPR will become available in a sustained time-frame that is compatible with its use as a maintenance P fertiliser. At say $350/tonne for an RPR/S mix containing 11.0% P and 8% S (the agronomically correct ratio), with $120/t transport and spreading this gives a cost of $4.27. This is a difference of about $10,000 on a 300 ha hill-country farm.
On the subject of RPR, Quinfert Algerian RPR is rated as a highly reactive phosphate rock suitable for direct application by the International Fertilizer Development Center (IFDC) in Muscle Shoals, Alabama, USA, the world authority on direct use of phosphate rocks as fertilisers. The naturally-occuring presence in Algerian RPR of 3-7% natural phosphatic dolomite (average 5%) can mean that the very shor (30 minute) citric solubility can be below 30% in some high-dolomite samples, although the eposit average is 30-30.5%. This is simply because the dolomite consumes a lot of the citric acid , leaving less to dissolve the RPR. It is a laboratory artefact: it is extremely important to note that this has no adverse effect whatsoever on its excellent field performance. I have a great deal of experience and a proven track record in RPR. The important question is whether the ‘RPR’ you are being offered contains all the P in the form of an identified RPR, or whether it is an RPR mixed with a non-RPR phosphate rock after arrival in New Zealand.
I am very sure that if I advertised a product as ‘super’ that was actually a mix of super and RPR, there would be hell to pay. Shouldn’t honest advertising work for everybody?
The comments and questions below from Robin Boom, agronomic Advisor, and the reply from Dr Quin, have been copied over from Dr Quin’s LinkedIn page (Bert F. Quin):
Comments and questions from Robin Boom,
Business owner at Agronomic Advisory Services, 8/9/2018.
Bert, I am of the understanding that Ravensdown are blending Moroccan BG4 with their Sechura to get the cadmium levels below 280 ppm/kg P whereas Ballance are mixing an Algerian based rock they call PB3 with their Sechura. With the new Biogro standards of not having RPRs with Cd levels above 150 ppm/kg P the ratio of the lower cit sol products increases. Your argument however about the dolomitic content of your Algerian rock lowering its 30 min cit sol, would also apply to Chatham Rise with its high CaCO3 content? Can you point us to the trial work with your Algerian rock? Is it the same product that Anton in Australia and Steve at Hawkes Bay and Shane have been marketing here in NZ for years? I have seen some of their claimed trial work done in Australia and it all looked a little dubious to me. For my RPR clients I am currently only comfortable recommending Ron Webby’s Sechura (45% Cit Sol) or Dickie Direct’s granular RPR (mid 30’s Cit Sol). Guano is fine but way too expensive.
Don’t expect Ballance or Ravensdown to respond to your critique of their superphosphate products and the available P and S content.
Response from Bert F. Quin
Managing Director at Quin Environmentals (NZ) Ltd, importer and distributor of QUINFERT Algerian RPR:
Yes, it is my understanding like you that they are blending beneficiated Sechura RPR (which contains about 12.7% P but averages up to 280mg Cd per kg P) with Moroccan BG4 grade phosphate rock from the Gantour region. This contains 13.0% P and low Cd, but it is NOT an RPR, and is not recognised by the industry-independent International Fertilizer Development Centre (IFDC) in Alabama, USA. Even the Moroccan’s themselves refer to it only as a DAPR.
I have been advised in writing by Ballance Agri-Nutrients Ltd that the product they label as ‘High-P RPR’ is a mix of Sechura RPR which has been beneficiated to ‘over 12% P’ , and a phosphate rock from Morocco known in the mining industry known as ‘PB3’, which contains approximately 14% P. My understanding from enquiries I have now been able to make from other sources is that ‘PB’ stands for PhosBoucraa, which is the mining entity in the disputed Western Sahara which mines the Boucraa phosphate rock. the existence of any field trial data from trials done with PB3.
he deposit is predominantly a deep (30m) layer which has a thin 1-3m vein of phosphatic dolomite running through it. The RPR contains 13-20 ppm Cd, average 18, or 140 mg Cd/kg P. Depending on the position of the phosphatic dolomite layer during a day’s mining (50-100,000t), the dolomite content can range from 3-7% of the total. There are very few other impurities present, unlike most other RPRs. The level of dolomite only alters the P content of RPR by +/-0.2%P (range 12-5-12.9, average 12.7%P); it is a high-P dolomite anyway. The presence of natural lime or dolomite in RPRs has long been known – in NZ and elsewhere – to give an artificially low reading in NZ’s very short (30 min) citric test. The citric acid preferentially attacks the more soluble dolomite, leaving far less citric acid to dissolve the RPR. Any chemist knows this. The pH of the citric acid solution when testing Algerian RPR increases in the first minute from 2.3 to 3.3. That may not seem to much to some, but you have to realise that the pH is a logarithmic (log for short) scale. At pH 3.3, there is only 10% or so as much citric acid to dissolve the RPR as there are at 2.3. So, depending how much phosphatic dolomite is present in a given sample, the citric solubility in the short 30 minute test can range from 28-37% , with a mean of 30-30.5%. The actual number has no effect on the effectiveness of the RPR as a fertiliser. These type of effects are detailed in several independent in scientific papers which anyone can find on google. An extreme case of this effect is found with the Chatham Rise Phosphorite (CRP) nodules (which Chris Castle hopes to mine). CRP contains 30% calcite (lime), which reduces the 30-min citric P to only 15%! That’s way lower than most phosphate rocks that be used for manufacturing only! Nevertheless, in every single field trial, CRP performed at least as well as any other RPR, right from the start. So you get my point hopefully. The 30-minute citric P test is simply not up to the job. As recommended by Dr Alec McKay (and myself for that matter) a far better test is do 5 sequential extractions. All the P will be extracted from an RPR, regardless of the dolomite or lime present, but not from a non-RPR. Finally, the initial high citric solubility of Sechura is itself a known chemical artefact. It does not mean it performs better in the field than other RPRs; it doesn’t, except briefly in some very low soil P situations.. Most unfortunately, this high solubility is abused by some players to get better citric solubilities for mixtures of it with unreactive rocks. The current 30-minute citric test was certainly better than nothing in the somewhat less devious 1970s, but is way, way past its use-by date now. A combination of the 5-sequential citric and the IFDC a-axis determination test need to be adopted by Fertmark without further delay.
History of Algerian RPR in NZ and Australia. I had been trying to get access to Algerian RPR since the earliest days of Summit-Quinphos. But Sumitomo, who owned about 80% of S-Q, would not let me go there because of the risk of kidnapping of visiting businessmen at the time. But I made sure I kept contact with the Algerian RPR producers (Ferphos’s subsidiary Somiphos) at IFA fertiliser industry conferences and by email. Ballance bought into SQ in 2000; I left in 2005. Sumitomo sold the company to Ballance about 2009. They first changed the name to ‘Altum’ , then completely merged the operation into Ballance, and gradually removed any advisory promotion of RPR.
About 2011 Anton Barton of BioAg Australia rang me for advice; he had been importing the very dusty Egyptian Red Sea Coast Kosseir RPR into Geelong, but it had been banned by the Port because of the dust. I set him up with the international trading company who have the agency for Algerian RPR in Australia, and he has been importing a bulk shipment of it into Geelong every 12 or 18 months since. I bought a few containers of it off him into NZ, as have Fert Wholesale Direct Ltd and one or two other companies. However, this is barely economic given the very high Tasman container rates. In 2017 I was approached (at the request of the Algerians) by the international shipping and trading company Agrifields, based in Dubai, who now have the shipping agency for Algerian RPR for all of south-east Asia and New Zealand. They asked me to market Algerian RPR in NZ. We agreed on terms, and I started importing containers direct into NZ in autumn. Somiphos make 2 grades; one specifically for direct application called 63-65, and another coarser grade for manufacturing (after very fine grinding). Provided the coarsest fraction is removed (and added back in after grinding if wished), this grade is also a perfectly good RPR, as it comes from the same deposit. Interestingly, a few hundred tonnes into NZ for superphosphate manufacturing trials about 2015, but it would have had a high sulphuric acid consumption because of its high liming equivalent (which as I mentioned is added to by the 3-7% dolomite content). It may well have been hoped it would be a good manufacturing rock that could also be pressed into service as an RPR if another RPR importer came along. The Algerian’s probably saw what was going on and asked me to come back into the market with bulk Algerian RPR, so I have!
Both the economic and environmental benefits of using RPR as the source of phosphate on New Zealand’s soils have been deliberately understated by the fertiliser manufacturing industry in the past decade. With no significant independent importation and scientific promotion of RPR, farmers who want to keep using it have found it increasingly difficult to access true RPR at a reasonable cost.
Three years ago, the industry was caught out selling substandard material as RPR, and now describe those products as DAPR (Direct Application Phosphate Rock) instead. The only true RPR that is being imported by the superphosphate manufacturers – Sechura RPR from Peru – has an undesirably high cadmium (Cd) content. And there are increasing reports that the unacceptable practice of blending this with cheap, lower-Cd manufacturing rock is continuing.
Why is this happening? Why are not greater efforts being made to supply true RPR to NZ farmers? The industry itself accepts that superphosphate manufacturing in New Zealand is on a slow but inevitable decline into oblivion, as they import more and more high analysis DAP and now MAP and even TSP to meet demand from their farmer shareholders. More and more of the sulphuric acid used to convert manufacturing phosphate rock into superphosphate is being imported directly rather than being made from imported elemental sulphur as traditionally done.
Even the argument that superphosphate has a ‘smaller carbon footprint’ than high analysis manufactured fertilisers does not stand up to scrutiny of the environmental emissions from superphosphate manufacturing, and the much higher costs, per unit P, of internal storage, freight and spreading costs with superphosphate. High analysis fertilisers containing 20-22% P require less than 40% of the tonnes to be stored, transported and spread and New Zealand.
So, given the increasing importation of high analysis fertiliser into NZ, why has the importation of true RPR with low Cd content virtually stopped? The reasons given – mainly poor supply – simply do not stack up. As an example, the industry says they will not import Algerian RPR – similar to Tunisian RPR but with lower cadmium – because it ‘does not meet NZ’s citric acid test’. This is incorrect; it demonstrates either a surprising lack of understanding of the deposit, or a deliberate attempt to confuse the farmer. More about this in a future Quinfacts. Other RPRs that could be imported are the Israeli Arad and the lower-cadmium end of the Tunisian deposit.
RPR has by far the lowest carbon footprint of any form of phosphate fertiliser there will ever be; it is simply dug out of the ground, crushed or washed and sieved to get rid of most of the accompanying clay and other minerals, and shipped here, to be mixed with elemental S and other nutrients and applied. Superphosphate served its purpose for 100 years building up available P levels in New Zealand soils, but they simply do not need to be increased anymore, just maintained where they are. RPR does that perfectly, as many, many trials have shown.
The science of the benefits of RPR
RPR has been researched extremely thoroughly in New Zealand. It will maintain pasture production and soil phosphate reserves at least as well as superphosphate in any situation except a combination of very low rainfall and higher than optimum pH. If annual fertiliser application has to be withheld, for example for farm income reasons, RPR maintains production far better than superphosphate. Sulphur requirements can be easily maintained by adding typically 7 kg elemental sulphur (either as fine S or sulphur-bentonite dispersable prills) for each 10 kg P required. Where soil sulphur levels have been allowed to run down to deficient levels (typically <5 ppm), the possibility of deficiency in early spring is easily avoided by the incorporation with the RPR/elemental S of any number of sources of sulphate-S, eg gypsum, sodium sulphate, sodium thiosulphate, calcium thiosulphate, ammonium sulphate and potassium sulphate, at about 10 kg S/ha.
RPR is a sandy mineral formed completely naturally by the decomposition of fish skeletons and shellfish on the sea floor, and their adsorption of phosphate from sea water, over hundreds of thousands of years. For cost reasons, almost all of the deposits being commercially mined are ones that have been raised above sea level by earthquakes or drops in sea level at some point in time. No chemicals are added to the product; beneficiation simply involves crushing and/or washing to remove impurities such as clay and organic matter. This increases the P content; this is further increased by drying to remove moisture. RPR is therefore a truly natural or ‘organic’ product, and is approved for use by organic farmers across the globe, as are gypsum and elemental S.
All true RPRs have an inherent liming effect of about 500kg lime per tonne of RPR applied. This is due to the fact that when RPR is applied to acid soil (as are all soils in NZ), the soil acid reacts with the RPR to release plant-available water-soluble phosphate. The soil acid consumed in the process reduces how much lime needs to be applied in the future to maintain the soil pH. This is particularly important in hill country where application costs of lime are very high. A typical soil in NZ needs about 250kg/ha lime annually to maintain the soil pH (this is typically done by applying 1 tonne every few years; this practice is simply to reduce the higher spreading costs of more frequent application of smaller amounts. The addition of elemental S reduces but does not eliminate the liming benefit of RPR. Long-term lime requirements are typically reduced by 30% where RPR/S is used to replace straight super, or 50% where it is used to replace sulphur-super.
On medium-high rainfall situations on low to medium P-retention soils, sulphur-super needs to be used instead of straight super. This is simply because the sulphate content in it is being leached too quickly to be of much use. As well as being economically wasteful (the farmer has paid for this sulphate), as it is leached from the soil it automatically takes with it ionically-equivalent quantities if calcium, magnesium, potassium and sodium. All of these unnecessarily leached nutrients ultimately need to be replaced, if production is to be maintained.
And finally, there is P run-off. A host of field studies in New Zealand and many other countries around the world have invariably shown that soluble P fertilisers such as superphosphate are far more prone to P run-off after rainfall events or irrigation than is RPR. This effect can remain for many weeks after application. The P lost in run-off enters streams, rivers and lakes where it is, alongside nitrate-nitrogen leached from cow urine patches, the biggest cause of eutrophication of waterbodies in New Zealand. Some recent industry-funded studies have argued that reducing ‘hotspots’ such as gateways and excessive soil P levels are more important than the form of P fertiliser used. However in my view, these studies are largely industry-backed smokescreens to try and hide the fact that if RPR was being used, these other loss mechanisms would simply not occur in the first place, or at least be greatly reduced. These hotspots are essentially caused by excess uptake of soluble P by pasture in the months after application. This excess uptake of P does not happen with RPR. More about this in a later Quinfacts.
Any phosphate rock will release some P into plant-available form when applied to acid soils. The question is, in the pastoral context, will it release P at a sufficiently fast rate to maintain the needs of a high production pasture? Most phosphate rocks do not. Even if very large quantities are applied they still do not, as the dissolution in the soil reaches equilibrium.
This is why, to be effective as a fertiliser, most phosphate rocks have to be chemically converted into a more soluble form of phosphate using a concentrated acid. The simplest process, to make single superphosphate, known as SSP or ‘super’, is to react the finely-ground phosphate rock with concentrated sulphuric acid. The acid reacts with the apatite tricalcium phosphate mineral which makes up the phosphate rock to form a mixture of water-soluble monocalcium phosphate (about 45% of the weight) and gypsum (about 55% by weight). Various greenhouse gases and others such as poisonous sulphur dioxide are emitted during this process. The product is then dried and granulated in most countries, or simply semi-granulated in the case of New Zealand, giving a product containing about 9% P. Typically a total of 8% P is plant available, with the rest being made up of unreacted phosphate rock and complex calcium-iron-aluminium phosphates. The gypsum provides about 11% S, all present as water-soluble sulphate-S.
If a wider ratio of sulphuric acid to phosphate rock is used, phosphoric acid is formed which can be separated from the gypsum. This phosphoric acid can then be reacted with more phosphate rock to form triple superphosphate (TSP, 20% P) or with ammonia to form diammonium phosphate (DAP, 18% N and 20% P) or monoammonium phosphate (MAP, 10% N and 22% P). These 3 ‘high-analysis’ fertilisers (and phosphoric acid itself), make up the vast majority of phosphate fertiliser traded internationally, because the high analysis greatly reduces transport, storage and spreading costs compared to SSP. SSP usage in NZ is slowly but surely giving way to imported TSP, DAP and MAP, amended with added elemental S to suit individual farm needs, rather than applying a blanket amount as with SSP.
It has however been known for nearly a century that some phosphate rocks can be used directly as a phosphate fertiliser on acid soils. All of New Zealand soils, and most of Australia’s agricultural soils, are acidic. The phosphate rocks that can be used with some level of effectiveness on acid soils are loosely defined as direct application phosphate rock or DAPR. With this number, there is a smaller group defined in New Zealand and Australia and an increasing number of other countries as reactive phosphate rock, or RPR. The fundamental definition of an RPR in the pastoral context is a phosphate rock that is reactive enough to release sufficient plant available P to maintain the growth of high-producing pastures. There are several laboratory methods used to help determine whether a DAPR is sufficiently reactive to be defined as an RPR. These methods are designed to avoid the need to do a whole series of new field trials whenever a new potential RPR is located or commercialised.
In the case of both New Zealand and Australia, nation-wide series of field trials comparing various DAPRs with superphosphate (SSP and TSP) were conducted in the 1980s and 1990s which ran for 4-6 years, followed in some cases by no-application for 1 or 2 years to measure the residual effects of the different fertilisers. Many other single-location trials and laboratory and glasshouse studies focusing on specific aspects were also conducted, and much research was conducted on predictive solubility tests.
The science of determining which phosphate rocks qualify as RPRs
In both New Zealand and Australia, it was determined that the simplest, reasonably accurate laboratory method to ensure that a particular DAPR would maintain pasture production as well as the same amount of P as soluble phosphate on acid soil, and therefore be described as an RPR, was whether or not at least 30% of the total P content of the DAPR would dissolve in a dilute organic acid (2% citric acid) in a short 30-minute extraction. RPRs that contained more than 30% ‘citric P’ or ‘citsol’ (as the test has come to be called) performed no better in this role as a maintenance P fertiliser, although they will give a slightly faster P response in very P-deficient situations. Although this citric P definition is not controlled legally, it is sufficiently strongly industry-recognised in both countries that using the term RPR to describe DAPRs that do not meet the 30% citric P test is unacceptable in the market, as both Ballance and Ravensdown have found to their cost in recent years.
When MAF analytical research chemists Mike Brown and myself first introduced the test back in the 1980s, it became rapidly accepted. A particular strength of the test was that, by doing the test on RPRs as sold, it provided an automatic control regarding whether the product as sold to farmers was fine enough. If the product was too coarse, the reduction in surface area meant that it did not dissolve fast enough – either to be fully effective as a fertiliser in the field, or to meet the 30% citric solubility requirement.
Over time however, two significant shortcomings in the citric acid test have become apparent. One of these produces an artificially low citric P for some true RPRs; the other allows one particular RPR (Sechura) to be blended with poor-quality phosphate rock and still produce a test of 30% or more. Let’s explain both problems.
- The effect of the presence of free lime or dolomite in RPRs. A small proportion of true RPRs contain a percentage of 10% or so free lime or dolomite mixed with the carbonoapatite mineral. Two examples are the small Hamrawein deposit on the Red Sea coast in Egypt and the massively larger Algerian RPR deposit. Much of the latter contains 10% free dolomite. This is even better from an agronomic point of view, as even less lime needs to be applied to maintain soil pH than with other RPRs. However, the presence of this dolomite selectively consumes about 20% of the citric acid in the laboratory test, leaving less of it to dissolve the P. So Algerian RPR, if mined where dolomite is present, can come up with a slightly lower citric P than 30%, even though it is just as good as a fertiliser. When applied to the soil, the individual grains of RPR and dolomite land in microscopically different locations, each consuming different micro-pools of soil acid. Fortunately, it is relatively easy to determine whether sufficient lime or dolomite is present to affect the citsol test.
- Sechura RPR has a different chemical composition in the crystal lattice to other RPRs, caused by the presence of hydroxide ion replacing some of the carbonate. This has no effect good or bad on its agronomic effectiveness – field trials have shown that it performs identically to other RPRs on pasture; but it does have a much higher citric solubility, ranging from 38-46%. Unfortunately, some unscrupulous companies have used this as a means to cheat farmers, by blending low-cost, low-solubility non-RPRs with Sechura, knowing that the average citsol can still meet 30%. Unfortunately for the farmer, the percentage of P present that is not an RPR will still be ineffective. One solution to this real problem is to insist that all companies selling RPR identify the source of their product, and sign a declaration that samples taken for Fertmark or other analysis contain only that RPR and no other source of P. It still requires honesty on the importer’s part however.
The a-axis measurement – unbeatable test for RPRs. I have been campaigning for 20 years for New Zealand and Australia to adopt an alternative RPR test, called the a-axis dimension test. It is a relatively simple test but requires a quite expensive piece of laboratory equipment – an x-ray diffractor or XRD.
The XRD measures the dimension in Angstroms of the a-axis of the carbonoapatite crystal lattice. RPRs have a smaller a-axis measurement due to carbonate being present instead of the phosphate in the lattice (note this is totally different carbonate to the carbonate in the form of free lime or dolomite present in some RPR deposits). Having a smaller a-axis makes the crystal less chemically stable and therefore more susceptible to being attacked by soil acid. The a-axis measurement that is used internationally to define the most reactive phosphate rocks, which is what NZ and Australia need, is a maximum 3.340 A (angstroms). This test cannot be fooled by a mix of Sechura RPR and a poor phosphate rock; the result will show two different peaks – one at 3.34 (Sechura) and one at say 3.37 for Moroccan. May not seem a big difference, but it is day and night for effectiveness. All the RPRs recognised in NZ have an a-axis measurement to larger than 3.340 A.
RPR is a natural, slow-release mineral formed on the sea floor over hundreds of thousands of years. Deposits that have been raised above sea level by changes in sea level or earthquakes are cheaper to mine.
When RPR is applied to acid soils, the soil acid attacks the phosphate mineral, releasing plant-available P in a sustained fashion. True RPRs release P fast enough to easily maintain the growth of high-producing pastures, with only a few limitations.
The science of using RPR in limiting situations
The first is that the soil has to have some level of acidity; otherwise the RPR will remain undissolved (as it has for eons in the North African desert). Because all pastoral soils in NZ and virtually all in Australia are acidic anyway, this isn’t an issue. It was originally advised that RPR should not be used at soil pH levels above 6.0, which is still slightly acidic (neutral pH is 7.0). The only reason for this recommendation was that RPR did not perform too well on a site that had a soil pH of 6.4. But this site also happened to have a very low rainfall (550mm). On farms with higher rainfall or irrigation, especially where existing soil P levels are already at or close to recommended maintenance levels, RPR has been shown to work perfectly well at soil pH levels up to 6.4. Few if any pastoral soils have been limed to a pH higher than this; it is simply not necessary for optimum production and induces trace element deficiencies.
The slow-release nature of RPR makes it less suitable where a rapid increase in soil P levels is required, eg in dairy conversions. This is particularly the case on high P retention soils.
Here are my recommendations for using RPR in less than perfect situations –
- Capital applications in very low soil P levels (less than two-thirds of the bottom of the recommended range):
Use a mix of RPR and a low-gypsum (non-superphosphate) soluble P, eg TSP, DAP or TSP. The water-soluble component should make up at least one half of the total P applied, until soil P levels reach the recommended range.
- Very low rainfall (<700 mm, no irrigation):
Following the end of a drought, apply the first P application as TSP, DAP or MAP.
- Soil pH in the range 6.4-6.7:
Use a mix of RPR and TSP, DAP or MAP. The water-soluble component should make up at least one-half of the total P applied.
- Very high P retention soils (PR>96%).
Use a mix of RPR with TSP, DAP or MAP for the first 5 years. The water-soluble component should make up at least 33% of the total applied P.
The cadmium (Cd) issue arose in New Zealand and Australia purely and simply because the manufacturing phosphate rocks both countries originally used to make superphosphate (Nauru and Christmas Islands) contained elevated levels (50-100 ppm) of Cd. All of this Cd ended up in the superphosphate, and therefore in the soil. Cd exists at low levels in all soils. But if too much is present in the fertiliser, the levels in the soil increase over decades to the point where elevated levels show up in pasture and then in animals, which concentrate it in their liver and kidneys, as happens to many poisons that cannot be excreted in the urine. The highest levels are found in soils which have had the highest application rates of superphosphate for the longest period of time; these are typically the dairy farms in the Waikato.
The highest levels of Cd are found in the liver and kidneys of young stock. To avoid the entry of excessive Cd into the human food chain, offal from young animals is not permitted to be sold for human consumption.
The reasons why some phosphate rocks are higher in Cd than others is not completely understood. Different hard ‘manufacturing’ phosphate rock deposits and different RPR deposits can contain very different levels of Cd. There can even be wide variation in levels within a particular deposit (more about this later). The concentration of Cd is thought to be largely due to the level of Cd in whatever organic matter is present in the phosphate rock as mined; that is, little is present in the phosphate rock crystal lattice itself. For this reason, the Cd can be removed by roasting the phosphate rock at high temperature. The organic matter in the phosphate rock is burnt off, along with the Cd in it. Ideally this is captured from the emissions. This roasting is expensive, and it also reduces the solubility of the phosphate rock, so it is not so suited for deposits that are used for direct application.
Internationally, most high Cd phosphate rocks (even those that otherwise are suitable for direct application) are used to make phosphoric acid. The reason for this is that most of the Cd precipitates out of the phosphoric acid and ends up in the gypsum. The phosphate fertiliser made from this phosphoric acid generally contains less than 20% of the amount of Cd originally present, but the enriched level in the gypsum by-product reduces its suitability for land application purposes, such as improving soil drainage and reducing salinity.
The science – and pseudo-science – of how Cd is managed by the industry in NZ
I first brought the issue of Cd accumulation to NZ farmers in the 1990s. Quinphos Fertilisers initiated a self-imposed limit of 2 ppm Cd for each % P in a product. This meant that an RPR containing 13% P could not exceed 2×13 = 26 ppm Cd. I based this limit on some excellent published independent scientific research conducted by Massey University, which showed that, at this level of Cd, there was no accumulation of Cd in the soil, and therefore in the food chain. At the time, we were importing Egyptian ‘Kossier’ RPR which contained only 12 ppm Cd, and Tunisian RPR, which ranges from 20-25ppm at the western end of the deposit to about 80ppm at the eastern end. We insisted that we receive RPR only from the low-Cd end; none of our shipments exceeded 25ppm. Averaged over the two RPRs, we had 19 ppm.
After years of ‘deny, deny’ that there was even an issue, the industry very slowly moved to adopting a self-imposed Cd limit of 280 micrograms of Cd per kg of P in any phosphate fertiliser. This equates to a 25 ppm Cd limit for a 9% P super, or 36 ppm for an RPR containing 13% P, nearly 40% higher than Quinphos’ maximum limit. This just isn’t good enough.
A decade ago the industry in effect admitted the seriousness of the Cd situation it had got NZ farmers into by introducing a ‘4-class’ categorisation of Cd in NZ soils. Farmers whose soils were in the highest Cd category could now only use fertiliser containing almost zero Cd.
What are the implications for RPR users? There are no RPRs with absolutely zero Cd (a cynic would say that the only motivation for having this pointless category at all was to prevent RPR being used). The main RPR now being imported by the industry (Sechura) is right at, or above, the industry’s own limit of 36 ppm. I have visited the Sechura deposit in Peru (as well as virtually all the other RPR deposits), as part of consultancy work for an international mining company. Unlike say the Algerian RPR deposit, which is essentially one thick homogeneous layer all containing 16-20 ppm Cd, Sechura has 8 different layers separated by clay; these layers range from 20-60 ppm Cd, with the lowest Cd layers being at the bottom. It is simply not possible to selectively mine the lowest Cd layers without enormous cost increase. The minimum weighted average achievable with current mining methods is about 38 ppm in my assessment. An importer may fluke one shipment at say 32 ppm, but the next is just as likely to be 45 ppm.
The very good news is that there are plentiful supplies of low Cd RPR. The Arad deposit in Israel is one. The Red Sea coast RPR deposits in Egypt are another. NZ’s underwater Chatham Rise phosphorite nodules – not mined as yet – are another. By far the biggest known to date, and already being mined, is the huge Algerian RPR deposit. All these RPRs are sufficiently low in Cd to lead to a slow but sure decline in the Cd levels in NZ and Australian soils if used as the source of phosphate.
Pasture plants need to take up 7 kg of sulphur (S) for every 10 kg of phosphorus (P). So, if both nutrients are being applied in reasonably efficient forms, a fertiliser containing 9% P should only need to contain 9×0.7 = 6.3% S. Near the coast, where there is significant input of sulphate in rainwater, the amount of S required is less or even zero.
But most people know that single superphosphate or ‘super’ contains 9% P and a fixed 11% S, almost twice the calculated amount of S required. Furthermore, in higher rainfall areas and low-medium P retention soils, even this amount is insufficient, and super has to be amended with elemental S to make sulphur-super. Why is this? It is simply because much of the sulphate-S gets leached from the soil before it can be used by the pasture (taking valuable cations such as calcium, magnesium and potassium with it). Even though 10-35% of applied soluble P becomes fixed in progressively unavailable forms in our soils, this is still more efficient than is sulphate-S, except in very dry conditions.
The science of S fertilisation
It has been known for decades that, except in very dry conditions and on very high P retention soils, finely ground elemental S is far more efficient than sulphate -S. This is why elemental S rather than say gypsum (calcium sulphate) is added to straight super if more S is required, and elemental S is the ‘go-to’ product for adding to the likes of TSP, DAP and MAP.
It is no surprise therefore that by far the most common blend of RPR and elemental S sold by Quinphos was 92% RPR and 8% fine elemental S. With 12.7% P in the RPR, which also contained about 0.8% S as sulphate, this gave a product containing 8.7% S and 11.7% P ; an S:P ratio of – you guessed it – 0.7 to one.
A big proviso is that the elemental S used has to be truly fine – mainly less the 250 microns and all less than 500 mm. This can be irritating to the eyes if not dampened. These days, the cost of prilling molten sulphur with bentonite clay is far less than it used to be. These small water-dispersable hemi-spherical prills contain hundreds if tiny sulphur particles in each prill, which disperse easily in the soil.
The only situations where sulphate is likely to be needed are on extremely low rainfall areas or in cold late winter/early spring if soil S levels are low. Sulphate of ammonia provides a good N boost and takes care of the S requirement in the latter case.