In this article, which first appears in a series on the global trade in water-soluble fertilizers, we look at two workhorses of this dynamic fertilizer sector – potassium nitrate and calcium nitrate. Looking at trade data is one of the ways to see new trends in consumption – which regions and countries are importing more of a product. But care has to be taken with the numbers.
Editor-in-Chief Luke Hutson explains.
This report will look at import and export numbers for 2019 for potassium nitrate (PN) and calcium nitrate (CN) available on the ITC portal. The aim is to give a general picture, provide some insight on the trade flows, and then highlight why caution is required with some of the numbers. The report will also refer to a trade matrix for both PN and CN generated by New AG International (NAI). Trade matrices are a vital tool in providing a snapshot for a given year, showing how much each supplier shipped to each buyer or destination. A trade matrix provides the detail on aggregate export numbers, and this can lead to insights on which markets are growing.
General points Some general points need to be made about these particular fertilizers at the outset. PN and CN would usually come under the umbrella term “specialty fertilizers” and this would imply that their volumes are smaller than bulk commodity fertilizers. PN and CN tend to be shipped in bags, and very large amounts of CN are also shipped as solutions, and in dry bulk bags of PN and CN. These bags are shipped in containers, hence the image used on the front cover. They are high-value products and bagging also counters the problem of adulteration.
PN and CN are often used on high-value cash crops, such as kale
Trade figures are trying to capture a snapshot of something that is inherently moving.
For both fertilizers, production is dominated by only a few suppliers. We refer to this as concentrated supply, which reflects the supply of potash as well. As will become apparent, Chile/SQM dominates the production and export of PN; Norway/YARA the production and export of CN. Having a concentrated supply does have benefits from an analytical view point. It can make it easier to deduce consumption figures – since many countries do not have any production of their own, the level of imports will also serve as an indication of the consumption minus any exports they make. And yes, even though a country has no production, it can still make exports, which means it re-exports some of its imports. It is worth pointing out for those who have not worked with trade figures before that export and import numbers invariably will not match. This can be a shock to the uninitiated, particularly in this digital age - shouldn’t these numbers match perfectly? The point that needs to be highlighted here is that any trade figures are trying to capture a snapshot of something that is inherently moving. Exports made in December might not arrive until January or February in the following year. So, exports for December could look higher than imports for December, for example. There can be other reasons – the product might be on a vessel making multiple discharges, such as three discharges at ports in three separate countries. Volumes to be discharged might change while the ship is in transit, or the final destination might change. For these reasons, it is vital to make what is called a trade matrix, which we describe as “reconciled” trade data. This is where the exports that country A says it has exported to country B are equal to the imports that country B says it received from country A. Typically, you would backfill a historical series with reconciled data. That said, this doesn’t necessarily mean a trade matrix is the final word and testament on trade, more that this is the trade that can be accounted for from an exporter and importer perspective. Consumption target The next general point to be made is that trade doesn’t tell you much about consumption, and knowing consumption is really the main analytical objective. In the simplest case, if a country has no production, then imports will normally equal consumption. But the problem comes when you have a country that has its own production of a particular fertilizer that exports that fertilizer and also imports that fertilizer, and when it uses some of that fertilizer to make other fertilizers that it then exports. (Hopefully, you are still with me!) And that’s before you even start looking at stock levels. Trade provides one piece of the jigsaw. New AG International has built production, import, export, consumption (PIEC) files for both PN and CN. The aim for this report is to bring out some highlights from the trade data, not to dissect the whole PIEC. Once you start looking at PIEC you need to go further and separate capacity and production. Some plants will be running at full capacity, and there are economic limits here too. You are unlikely to run a plant below a certain percentage of its capacity before it becomes uneconomical. Of course, it depends on what type of plant you are talking about. Nitrogen plants would be run at high levels, >80 percent, but a potash or phosphate mine can be run at a lower level. For sulphur, largely a biproduct of the refining industry, you are talking about the refinery operating rates, so an indirect measure driven by demand for fuel, not even an agricultural commodity. The utilization rate of a plant or factory would again form part of a PIEC spreadsheet. Actual production volumes are essential to know, rather than
capacity numbers. In an ideal scenario, if you knew the actual production of every single producer, you could assume this equates to global consumption (putting aside the question of stock building). But there is no such perfect knowledge. You can make assumptions on how many days per year a plant will run, but what happens about that maintenance turnaround that was extended by a few weeks – what did that do to production? But in essence, if we assume production and consumption should roughly be the same, then trade is what happens in the middle. This is where the product moves from geographical regions of surplus to regions of deficit. Apparent consumption In the ideal world that was mentioned above, it would be nice to have actual fertilizer consumption figures published by each country. Some do – for the main straight fertilizers. Some publish consumption figures at the nutrient level. Some go by delivery records, so it might still be sitting on the farm. The work-around is apparent consumption, a derived number. If you go to a textbook, you’ll find the following formula:
For some countries, where they published consumption, it can even be used to derive an apparent production figure if necessary.
And then there is the problem mentioned above, where countries have no production but are reported as having made exports, often called re-exports. Often when looking at trade data you will see that a country has some exports, but you know it has no production. One possible reason is when a destination is landlocked. South Africa often shows exports for fertilizers that it doesn’t produce, but it is the point of entry for southern Africa. A country being the point of shipment not origin is often seen in the Baltic ports. Often fertilizer is produced in Russia and then railed to a Baltic port and, for example, shipped from the port of Muuga in Estonia. Exports from Estonia are likely to be Russian, so this is another reason for having a trade matrix. You don’t really want to put down exports for Estonia when these aren’t really exports from Estonia. You really want to take them off Russian production to give a more reliable figure for Russia’s consumption. When devising a PIEC, as a general rule of thumb, you don’t want your overall global consumption figure running above your global production number. It is possible there was a build-up in stock for a few years, and so consumption could be higher than production, but this is unlikely to persist long term. The other problem with consumption is stock levels and time delays. A country might report a jump in imports in one year – making analysts jump with excitement that this was a growing market. But that doesn’t mean it was necessarily all consumed in that year, and it might have been
Apparent consumption = production – exports + imports – change in stocks
re-exported if there was a commercial opportunity. Raw material problem With PN, there is also the problem that it is included in NPKs and exported in that form. This is the case with Haifa in Israel, where it exports PN and uses some for its NPK products which it also exports. If you only took the PN exports, this would give Israel a higher PN consumption than it really has. This is another reason for constructing a trade matrix for each specific year to see detail of where exports are going, and then seeing what remains of an estimated production. In the case of Haifa, it is necessary to include PN into the NPK exports. Exporter information Exporter information – namely the list of countries that an exporting country says it exported to – can also shed light on the demand from that destination, particularly if that country does not report its PN imports. When constructing the PN trade matrix, exporters reported India as a destination for around 20,000 tonnes (t). There is no clearly published capacity for PN in India (though some may exist), so once some small volumes of exports are deducted, it implies consumption could be 18,000 to 20,000 t. Technical end-use Other factors for both PN and CN is that not all production is for agricultural end use, and is destined for what is called industrial use, such as solar concentrated power storage for PN and explosives for CN (and PN). PN is used in glass making and as a food preservative. And even with agriculture, the agricultural end use might only partly be nutritional – some could be to induce flowering, such as the usage of PN in the Philippines. The industrial component needs to be broken out in the numbers, but this is never straightforward. Ultimately, it needs to be done on a country by country basis. At some point, you will be hoping to find a figure for that country for agricultural consumption, and then apportion the balance to industrial usage. In China, one published source for PN consumption put industrial usage as high as 50 percent. Potassium Nitrate Potassium nitrate (PN) (NOP, 13-0-46) is the main water-soluble straight fertilizer, and the primary ingredient in most water-soluble NPKs. NOP is dominant in the world of water-soluble fertilizers (WSF), because in the same molecule it combines macronutrients nitrogen and potassium, consumed by plants at high rates. The importance of PN as a WSF means, in general terms, the largest producers have also been the world leaders in the market for soluble fertilizer, including finished products. There are two main producers in the world – SQM in Chile and Haifa in Israel. This is reflected in the export figures (see Table 1.). Indeed, such is the dominance of Chile, that it accounts for just under half of world exports and has done so for many years. That appears to be the case in 2019 when Chile’s PN exports were reported at 461,000 product t. The producer in Jordan is Kemapco, and in China there are several producers of which Migao Corporation is probably the best known. Other lower production capacity is located in Spain, Belarus, Russia, and Ukraine.
Potassium nitrate salts are used as a thermal battery for concentrated solar power (CSP) storage. The Andalusian Spanish complex Andasol was Europe’s first parabolic concentrated solar power station, built in 2008.
Port of Rotterdam, Netherlands
Trade matrix From the NAI trade matrix for 2019 using ITC data, global trade for PN is estimated at 1,005,000 t, which would represent 32 percent of estimated production. In many ways we would expect that percentage to be high because of the concentrated supply, and so product needs to move to areas of demand. If you compared this with ammonium nitrate trade as a percentage of production, it would be around 20 percent and for urea it would be 28 percent. One reason it is not higher is the dominance of Chinese production and consumption. This reduces trade as percentage of production. If you were to remove Chinese production and exports, then trade as percentage of global production increases to 61 percent. At this stage we just need to be wary since this is not all for agricultural use. Our adjustment factor will be something like 70 percent for agriculture (discussed below). Only further analysis, country-by-country, will refine this number. PN importers The following Table 1 shows the top 10 PN importers from ITC data for 2019. These top 10 countries account for around 74 percent of total imports for 2019. There are 80 countries reporting imports in the ITC data, which suggests many import small quantities. This would be expected for a specialty fertilizer, where cargoes can be a few hundred tonnes in shipping containers. With this table, you basically have two tiers. Spain, Netherlands and the U.S. out in front, with Turkey some
Data: ITC 2019
way between the others. With the Netherlands, care always has to be taken with any trade statistics because of the Rotterdam effect – this is where exports from and imports to Rotterdam often get coded – incorrectly – as exports or imports for Netherlands. PN Exporters Potassium nitrate exports are dominated by Chile. The domestic market is around 60,000 t (See New Ag International June 2020). The 2019 trade matrix shows exports of 460,975 t from Chile, followed by Israel, China and Jordan. There is a big drop to Spain.
PN consumption Global apparent consumption according to the generated PIEC using ITC data is estimated at three million t (2.996 million t) for 2019.
Given the comments made at the start, consumption numbers often need to be derived in the form of apparent consumption. The data will be presented in later reports. But just to highlight one country to show the steps necessary to disentangle the agricultural and industrial usage, Australia is a good example – it has a well-known mining sector and it has a growing greenhouse sector. The country has an estimated apparent consumption of 20,000-25,000 t PN, given imports from the trade matrix of 20,679 t. It is likely some of that would be used for explosives given the country’s mining industry. Using a calculation shown in references (Ref 1) based on data from ABARES, it appears that half of this PN volume is probably for agriculture and the remainder for industrial.
Calcium nitrate Calcium nitrate (CN) (15.5-0-0+26.5CaO) has a high solubility and is the third most import ingredient in the soluble straights market after potassium nitrate and technical-grade MAP. CN is the most important source of calcium in fertigation. This product can be described chemically in different ways depending on the hydration, which will in turn dictate its solubility. This has implications for the HS code, which is discussed below. In previous New AG International articles, we have estimated some 60 percent of the world’s CN is applied by fertigation and foliar feeding, while the balance is applied by dry application. Both the fertigation and foliar segments have been growing – our last figure was >5 percent per year – and there is an assumption this will continue, mainly due to growing acreage of fruits and vegetables. CN is not compatible with WSF phosphorus and sulphate raw materials, such as MAP, MKP, SOP, AS and Mg-sulphate, so very few CN manufacturers also produce NPK products. That said, a large majority of supply is produced by Yara – as its YaraLiva-Calcinit product. For Yara, CN is a byproduct in the production of phosphorus products. Other producers include Fertiberia in Spain (at the site of former Portuguese ADP), the Czech producers Lovochemie, Polish producer Adipol and ADOB, and various Chinese companies.
In 2016, Russia’s Uralchem presented a tech-grade anhydrous product (CN (17-0-0+33 CaO). Chemical formula and HS code As can be seen from Uralchem product, it has a slightly higher N content then what might traditionally be called CN. This raises the issue of how variations in chemical composition of CN reveal themselves in HS code, and as a result could lead to misalignments in trade data. There are two points to this discussion – the chemical formula of CN and then the HS code. Firstly, the chemical formula. The CN salt may occur as three hydrated salts, and an (anhydrate) anhydrous one.
Ca2(NO3)2*2H2O, Ca2(NO3)2*3H2O, Ca2(NO3)2*4H2O.
The latter with 4H2O is known as tetrahydrate and represents stable solid phase at room temperature. The tetrahydrate seems to be the standard reference, but one large producer gives its CN formula as Ca2(NO3)2*2.5 H2O. An (anhydrate) anhydrous product would have the chemical formula Ca2(NO3)2. When discussing this issue with Dr. Oded Achilea, NAI’s contributing editor, it would seem the ideal is to have as little crystallization water as possible, which means that product is higher in specific nutrient content, such as calcium in this case, which means the producer could command a higher price. But less crystallization means higher hygroscopicity which can lead to other problems. One word of warning – sometimes a product can be called calcium nitrate but it is calcium ammonium nitrate when you look at its chemical formula under the product specifications. Ca2(NO3)2*NH4NO3*10H2O. There is also a distinction between liquid and solid forms of CN. When talking to a fertilizer producer, they said anhydrous CN also had a lower N in NH4 form content than standard soluble grade. The level of NH4 is one reason why a grower might use a liquid form of CN which might be completely free of NH4.
Now on to the HS codes, which is how you identify a product in international trade. For CN there are two codes: HS 283429 and HS 310260 – • 310260 Double salts and mixtures of calcium nitrate and ammonium nitrate • 283429 Nitrates excluding of potassium and of mercury If you look up the HS code for a given CAS number – then the CAS number for CN tetrahydrate will return the HS code 283429. If you put in the CAS number for CAN you will get 310260, which is also 310240 (mixtures of AN with calcium carbonate). Therefore, 283429 has been used in the following data. Just taking one example, if you use 310260 code you see China exporting 533,000 t in 2019 which is more than its 255,000 t CN capacity (albeit it's estimated capacity, so it might be higher). So, what else could be included in this 310260 code? The likelihood is this volume contains AN products. China does not export CAN.
CN trade CN provides a good example of the disparity that can exist when looking at total import numbers and then looking at a trade matrix. Global imports for 2019, according to ITC data, were 1.9 million t. But the trade figure from the NAI trade matrix is 629,000 t in 2019, using the same ITC data. We’ll explain the discrepancy below, but essentially Norway does not report its exports, so only countries reporting imports from Norway are included, and in the trade matrix only 220,000 t could be found from Norway, yet production from Yara is more than one million t. Even adjusting for that will not get the trade matrix figure to 1.9 million t. With CN, there are similar problems to PN, in separating out the technical and agricultural uses. CN is also used in explosives and used in the construction industry to make concrete. Perhaps a lesser known application is its usage in the production of latex gloves. Given the Covid-19 pandemic, we might see more demand for CN for glove production, although this would not translate into a demand increase of large volumes. CN importers Global CN imports were 1.9 million t in 2019 according to ITC data. Top importers are detailed in Table 3. That is a close grouping for a top 10 of importers, compared to PN which dropped off quickly from the top three.
But we need to be mindful that not all imports are accounted for – this is the value of doing a trade matrix. As we’ll see in the next paragraphs, we need to take into account the Norway problem. CN Exporters One of the key decisions is how to treat Norway exports. Norway doesn’t report exports of CN. This is when we go back to the Yara financial reports and for 2019 see total CN production of 1.542 million t, of which 400,000 is reported as technical. We can use this for two purposes – the first is to estimate the Norwegian CN exports. Let’s say some of the CN is used for other products. One source to NAI has estimated the volume at 700,000 t from Norway, around 500,000 t more than that found in trade matrix. This still seems on the low side. If we add an additional 500,000 t to the trade
Table 3 - CN Importers Top Ten (product tonnes)
matrix figure, we arrive at trade of 1.1 million t, which is still short of the 1.9 million t of total imports from ITC reporting countries. This implies Norway’s exports are closer to that production figure. If we assume exports of 1.2 million t, which allows a volume into other products, this brings the trade matrix towards that 1.9 million t figure. The Yara figure can also be used to refine the estimate for the agriculture/industrial split. From the Yara figure, approximately 30 percent is listed as technical. And so our working assumption, which we’ll extend to the global split, is 70 percent of consumption is for agriculture. China is the second largest exporter with 185,000 t from the trade matrix. From a domestic production of around 255,000 t this suggests an apparent consumption of around 75,000 t. That consumption is less than the U.S. China’s reported exports for 2019 were 198,000 t but only 185,000 t could be found for the trade matrix. China’s consumption is less than the U.S., does this seem reasonable? The production capacity might be understated. The NAI capacity is separated by company, and so this is more solid than simply being given an aggregate figure for China capacity. But to export 70 percent of production seems high. One possible explanation is that the local market is being developed and eventually these exports will reduce. As can be seen from the following Table 4, Norway dominates exports, followed by China and then Poland, according to the ITC data.
* estimated by New AG International
Table 4 - CN Exporters Top Ten (product tonnes)
CN consumption Global apparent consumption is slightly less than PN, at 2.6 million product tonnes across agriculture and industrial uses, according to the NAI PIEC model, and fitting with published estimates of production capacities. It is worth remembering that even allowing for stock-building effects, consumption is unlikely to run above production capacity for any length of time. Using our estimated split of 70 percent agriculture, this suggests 1.8 million t CN for agricultural use worldwide. When trying to gauge the consumption level for the average sized market, we’re talking in the tens of thousands. The U.S. is the largest consumer, with a combination of domestic production (of around 60,000 t) and reported imports of around 80,000-90,000 t per year; less some exports, an apparent consumption of around 135,000-140,000 t. Egypt would be the second largest consumer because of a production capacity of 120,000 t. The country exports very little.
Another way that a trade matrix can be of value – few exports were picked up from Slovakia which, given the country has a production capacity of 80,000 t, would give an apparent consumption at a similar level to China, which wouldn’t make sense. Separate research suggests exports of 65,000 t and therefore a consumption of Slovakia of 15,000 t, which fits with our general assumption of consumption in the tens of thousands. ●
The following table summarizes where we have been in this first article in the series. We have seen the concentrated supply for both PN and CN and how this results in the traded volume forming a high percentage of apparent consumption. In the case of CN, the dominance of Yara production makes this percentage even higher. PN is also dominated by a single producer in Chile. Both products have agricultural and industrial usage, sometimes in small quantities, and this also explains why a high percentage is traded. But it presents a headache – how to estimate the split and therefore volumes involved in agricultural and industrial end use. ●
The following table offers a summary of the discussion in this analysis:
Table 5: Summary - Based on ITC 2019 data
Ref 1 PN Australia calculation based on a reference from here: ABARES fertilizer use year-end 2017 PN applied to 158,000 ha in Australia. Assuming a recommended rate for vegetables of 50-100 kg/ha, and using mid-point. 0.075*158,000 = 11,850 tonnes of PN for agricultural use. Thanks go to NAI contributing editor Dr. Oded Achilea for work on capacity and production tables.
Port of Shanghai, China
Bringing the latest trade figures into focus is New AG International’s Editor-in-Chief Luke Hutson.
