Formulation
The calculation of blend formulations is not a difficult process but it is an essential requirement for producing good quality blends. The process very often is performed by a computer and combined with weigher or metering controls. Records should be kept of each formulation and this can be simplified by the use of a standard format in the computer. For accurate formulation the actual raw material analyses should be used but standard analyses are helpful as a first approximation.
Example formulation to produce a 20:10:11 product from the following raw materials:

1 To obtain 11% K2O in the final product requires 18.5% Potassium Chloride.
2 To obtain 10% P2O5 in the final product requires 22% Di-ammonium Phosphate and this will provide 3.9% Nitrogen.
3 The remaining 16.1% Nitrogen in the final product will require 59.5% Calcium Ammonium Nitrate.
4 Total raw materials required adds up to 100% of the total space and thus no filler is required.
Formulation per tonne of product is thus:
595 kg of Calcium Ammonium Nitrate
220 kg of Di-ammonium Phosphate
185 kg of Potassium Chloride
NOTE: Formulations containing less than 5% of any ingredient should be avoided as there will be difficulties in mixing this small amount evenly throughout the mixture.
Screening
Raw materials should be screened before being fed to the blending unit to remove any lumps or fines which may have formed during storage. If screening is performed after weighing and blending, some of the raw material will be lost and this will alter the composition of the final mix. In such cases there can be no control on the final analysis of the blend. In addition, the screening process may seriously un-mix the components. The use of a scalping screen (for example 10mm mesh) to remove any lumps which may have formed in the system, is acceptable.
Sampling
Random spot checks of the raw materials being fed to the blender are recommended. These may be simple checks of the particle size distribution, but if there is any doubt representative samples should be taken and examined fully. Representative sampling of any material requires special techniques and equipment. Full details are outside the scope of this resource but it is recommended that blenders follow recognised sampling methods such as those set out in European Standard EN 1482 Part 1[1].
It is essential that all samples are correctly taken to ensure their representativity. There is no point in taking and analysing unrepresentative samples. The quality of the measurements made on the samples depends on their representativity. This is particularly important for blended fertilisers where the risk of heterogeneity is higher.
As stipulated by CEN, blended samples must always be taken from material in motion, either in free fall or on conveyor systems. It is essential to sample through the whole stream rather than from the same part of the stream. Suitable automated equipment is available for most circumstances and is to be recommended whenever possible.
For the blend producer it is important to know the sampling and measurement methods and the necessary equipment. Generally, all of this is described in standards which are regularly updated. Blenders are recommended to study the latest versions of sampling and measurement standards.
When a sample of any granular matter must be taken, some questions should be asked:
1 What quantity must I take?
2 How many increments to obtain this quantity?
3 How to take the samples?
Finally the sample must be reduced in order to obtain a final sample for analysis (usually about 250 g for a sieving test and 500 g for a chemical analysis). Thus after the sampling there is also a reduction step. Full details of reduction methods are given in EN 1482 Part 2[1].
Sampling Quantity
The minimum quantity of sample recommended is given in EN 1482 Part 1. The recommendations are summarised in Table 1.

If the weight of a bag is less than 5 kg, a complete bag is considered as a sub-sample. The number of selected bags can be higher if the final quantity of the sample is not sufficient. For bags of 50 kg, the content must ideally be divided in order to obtain a representative sample, using a sample divider. Sampling the contents of an IBC (big bag) is not easy. The best solution is to sample during emptying, but this is not always possible. Further information is given in EN 1482 Part 1.

The data from this table can be expressed in a more practical way (Table 3).

The amount of fertiliser for an increment is about 250 g for raw materials and 500 g for blends.
Sampling Equipment
Appropriate sampling equipment is essential. There are various automated systems for use with conveyor belts. These are certainly recommended if there are large amounts of fertiliser handled. For smaller deliveries, manual sampling is more suitable. A suitable type of sampling cup is described in EN 1482 Part 1[1] and is essential for sampling from falling streams of materials (Figure 1). Information about the number of increments to be taken must be given to the operator (Table 3).
Sample Divider
Rotary sample dividers are recommended, particularly when sieve analysis is to be carried out or when blends are being tested. Riffle dividers may be used for raw materials but are less suitable for blended products. Descriptions of these dividers (Figure 2) can be found in EN 1482 [9]. It is important to remember that for the measurement of the physical properties, the particles must not be crushed, but for chemical analysis, it is recommended to crush the particles before the final reduction of the sample.
Sample analysis
Chemical analysis
Standard reference test methods for most of the common chemical analyses used for fertiliser materials may be found in EC Directives, National Standards, European and International Standards and the AOAC Methods of Analysis Handbook. It is not essential that these methods are used for routine process control and simpler, often automated, methods are available. However, all such methods must be evaluated against one of the recognised standards.
Physical tests
Recognised standard test methods exist for several physical properties. The interest of blenders should focus on test sieving and bulk density. Some laboratories have an accreditation for these special measurements. No Standard methods are available for such properties as caking and free dust.
Sampling is also a fundamental step for physical testing because of the segregation that occurs naturally. This is particularly the case for sieve analysis because of the size segregation when there is a flow of fertiliser (which is always the case for bulk products). Another point is that for physical testing, the quantity needed for the final sample may be more than 5 kg. This means that the number or the mass of sub-samples must be sufficient to achieve this.
Test Method for Sieve Analysis
(The following method is a summary of EN 1235/A1)
Principle:
Dry sieving of a sample of fertiliser material with one or more test sieves, using a mechanical sieving machine.
Apparatus:
• Balance, capable of weighing to the nearest 0.1g.
• Stainless steel woven wire test sieves, 200mm diameter, with a lid and receiver for the sieves.
• Mechanical shaker (sieving machine), capable of imparting both horizontal and vertical motion to the material on the set of sieves.
• Stopwatch
• Soft brush
Procedure:
Reduce the sample to approximately 250g, preferably using a rotary sample divider, or if one is not available, a riffle divider. Select seven sieves to cover the range of particle size expected and assemble in ascending order of aperture size on top of the receiver. (See note at the bottom of this table)
Weigh the test portion to the nearest 0.1g and place it on the top sieve and fit the lid.
Place the set of sieves with the sample on the shaker and shake for 10 minutes.
Remove the sieves from the nest, starting from the top and weigh the quantity retained on each sieve and in the receiver, to the nearest 0.1g. Remove any particles trapped in the mesh by brushing from underneath.
Sum the masses of the fractions retained on the sieves and in the receiver and check that the total mass is within 2.5g of the original mass. Calculate each fraction mass as a percentage of the sum of the masses and draw up a table showing the cumulative percentage passing each sieve.
The percentage of material retained in the receiver (x0) and on each sieve (xn) is obtained from the formula :

Where:
mn is the mass on sieve n
mt is the total mass (m0 + m1 + …)
Xn is the mass % retained on sieve n
The cumulative undersize is defined by the formula: Cn = X0 + X1 + X2 + … + Xn-1
Where Cn is the cumulative % undersize for sieve n.
Test Method for Loose Bulk Density
(The following method is a summary of EN1236[2])
Principle:
Weighing a known volume of the fertiliser.
Apparatus:
• Balance, capable of weighing to the nearest 1g.
• Cylinder of a known volume, V (approx. 1 litre and diameter around 60 mm).
• Standard funnel with an aperture of 25 mm diameter, Figure 3.
Procedure:
Place the sample of fertiliser in the funnel with the aperture closed. Weigh the empty cylinder and place it under the funnel. Open the aperture of the funnel and allow the fertiliser to flow into the cylinder. When the cylinder is full, close the funnel and remove the excess fertiliser above the cylinder with a spatula. Weigh the cylinder and its contents and calculate the weight of fertiliser (m in kg).
The loose bulk density is given by the following formula:
Test Method for Angle of Repose

(The following method is a summary of EN 12047[3])
Principle:
Measurement of the diameter of a heap of a given height and calculation of the angle of the heap.
Apparatus:
• Funnel (aperture of 25 mm diameter) placed at 120 mm above a surface, Figure 4.
• Horizontal surface of 750 x 750 mm, with four lines presenting an angle of 45° between them and traced at the centre of the surface.
Procedure:
Place about 5 kg fertiliser in the funnel with the aperture closed. Open the aperture and allow the fertiliser to flow freely onto the surface and form a heap. The flow stops when the heap reaches the bottom of the funnel. Measure the four diameters on the plate. Calculate the average diameter (d in mm). The value of the angle of repose is obtained by the following formula:

Quality control
Blenders should draw up a routine quality control schedule to include the sampling and analysis of raw materials and products as well as checks on raw material weighers and/or feeders and check weighing of finished bagged products.
Raw materials
The amount of quality control of raw materials depends on the reliability of the suppliers. If supplies are received from single sources, experience will soon establish the degree of control needed. If raw materials are purchased from a variety of sources, extra control is essential.
Random samples from each consignment should be taken as described here, for reference purposes. These samples can be sent for analysis whenever there is reason for doubt but they should be carefully labelled with date and origin and kept for at least three months.
The particle size distribution of the raw materials is the most important property to be controlled. Representative samples of raw material feeds should be taken and tested as detailed here at regular intervals with a minimum frequency of once per shift. A full sieve analysis should be carried out on each raw material.
Finished products
All blenders have an obligation to meet the requirements of their National Fertiliser Regulations and for those fertilisers which are marketed as EC fertilisers, common Regulation EC 2003/2003 [4] applies throughout the Community. National Regulations will cover products which are not declared as EC fertilisers. It should be noted that in cases where the farmer does not purchase a fertiliser with a specified nutrient ratio, for example, if he purchases his own raw materials and contracts the blender to mix them for him, all official controls for the blended fertiliser may not apply.
The regulations set out the tolerances on the declared chemical analysis for all fertilisers. Fertilisers not meeting these tolerances may result in prosecution and penalty. It is therefore recommended that some degree of quality control is exercised by blenders (auto-control).
The amount of quality control required will depend very much on the scale of the blending operation. For simple small scale batch blending operations it may be sufficient to rely on the principle that “what goes in must come out”. However, even these operations will need some regular control of the weighing equipment. For large scale continuous operations, consideration should be given to full automated on-line sampling and analysis.
In between these two extremes, it is recommended that one representative random sample is taken from one of the grades made each day. If laboratory facilities are available on site, these samples should be analysed daily. If outside contract laboratory services are used, the frequency of analysis should be at least one sample per week, selected at random from the week’s collection.
Records should be kept of all samples taken and analyses carried out.
Micronutrients
It is relatively simple to add micronutrients to blended fertilisers either as granular or powder materials, but the homogeneity of distribution is particularly important. As a guide, granular materials should not be used if the micronutrient carrier represents less than 5% of the total weight. In these cases it is preferable to add the micronutrient in powder form with a binding agent to ensure good adherence to the granules. Suitable binding agents include heavy viscosity oils, UAN solutions and water. Alternatively, the micronutrients may be added as a solution which is sprayed directly onto the blend in the mixer. Note however, that oil should not be used when ammonium nitrate forms part of the formulation. Some national regulations do not allow the use of oil: check this point carefully.
Conditioning Agents
Anti-caking agents are not normally required if the product is to be used immediately. Products which are to be stored or bagged may require the addition of a small amount of anti-caking agent. It is recommended that a dust-suppressant is added to the blend. Conditioning agents may be applied in the same operation as the micronutrients.
References
1 CEN, European Standard EN 1482. Sampling Parts 1 & 2
2 CEN, European Standard EN 1236. Determination of bulk density (loose)
3 CEN, European Standard EN 12047. Measurement of static angle of repose
Links to related IFS Proceedings
533 (2004), Practical and Theoretical Aspects of Fertiliser Sampling, D Thompson
546, (2004), Micronutrient Inclusion in Fertilisers: Safety and Compatibility, H Kiiski
Need more information?
If the information you need on this topic is not on this page, use this button to access the relevant section of the forum, where this may have been provided. If not, you may ask the question.