STATISTICAL QUALITY
CONTROL OF
DIAMMONIUM PHOSPHATE
(DAP) FERTILIZER
Aditya Subramani Murugan, 1387981
Chandrasekar Ramasubramanian, 1363000
Duarakesh Sridhar, 1450764
Harihara Sekar, 1433826
Naveen Reddy Jakka, 1444409
Vegnesh Subbukrishna, 1440604
APRIL 14, 2016
INDE 6336: STATISTICAL PROCESS CONTROL
Department of Industrial Engineering
Table of Contents
Abstract: ........................................................................................................................................................ 2
Background: .................................................................................................................................................. 2
About Greenstar Fertilizers Limited: ........................................................................................................ 2
About Diammonium Phosphate fertilizer (DAP):..................................................................................... 2
Manufacturing Process: ............................................................................................................................ 2
Important parameters affecting quality: ........................................................................................................ 3
Sampling Procedure: ..................................................................................................................................... 4
Earlier method:.......................................................................................................................................... 4
New method: ............................................................................................................................................. 4
Pre-Neutralization process monitoring: ........................................................................................................ 5
Total Nitrogen Content: ............................................................................................................................ 5
Total Phosphate Content: .......................................................................................................................... 6
Process capability of pre-neutralizer: ........................................................................................................ 6
Granulation process monitoring: ................................................................................................................... 7
Further monitoring of particle size: .......................................................................................................... 9
Process capability of granulator: ............................................................................................................... 9
Conclusion: ................................................................................................................................................. 10
Results: .................................................................................................................................................... 10
Suggestions for further improvement: .................................................................................................... 10
References: .................................................................................................................................................. 10
Abstract:
Greenstar Fertilizers Limited (GSFL) is a leading manufacturer and marketer of fertilizers in India that
manufactures phosphatic fertilizers since 1978. One of its most reputed products is its Diammonium
Phosphate (DAP) fertilizer. DAP is used as complex fertilizer for supply of N and P2O5 nutrients to
plants. GSFL offers DAP at a price point higher than its competitors but provides superior quality than
them. Thats the reason why customers choose GSFL over its competitors. The most important
characteristics of DAP are its Total Nitrogen percent by weight, Moisture percent by weight and Water
soluble Phosphates (as P2O5) percent by weight. Control of these characteristics are dictated by government
regulations and the need to give a better product to the customers. A deviation from the standards will result
in the revoke of license from the government, destruction of soil quality and also the loss of customers. So
quality control is very important for them. But the number of complaints from the customers has been
increasing lately and they are moving to cheaper alternatives. The need for tighter quality control, better
sampling techniques and quicker testing methods are evident from the current scenario. In this project, we
analyze the efficiency of the current quality control, sampling and testing techniques used by GSFL, look
for any weaknesses in the system and provide suggestions to correct it. We believe that a sincere effort to
improve the quality control will result in better quality product and improved soil quality for the customers.
Background:
About Greenstar Fertilizers Limited:
Greenstar Fertilizers Limited is a leading manufacturer and marketer of fertilizers in India. The company
is also a brand launch platform for organic and inorganic fertilizers that stand up to its tough quality
certification process. Greenstar also imports fertilizers for sale in India. A public company since August
2011, Greenstar has acquired the phosphatics manufacturing assets of SPIC, an enduring fertilizer brand in
India, bringing it a significant strength in phosphatics.
About Diammonium Phosphate fertilizer (DAP):
Diammonium phosphate (DAP) is the worlds most widely used phosphorus fertilizer. Its made from two
common constituents in the fertilizer industry, Nitrogen and Phosphorous, and its relatively high nutrient
content and excellent physical properties make it a popular choice in farming and other industries.
Specifications:
Chemical Formulation (NH4)2HPO4
Moisture per cent by weight (Maximum) - 2.5%
Total Nitrogen per cent by weight (Minimum) - 18%
Ammoniacal nitrogen form per cent by weight (Minimum) - 15.5%
Total Nitrogen in the form of Urea per cent by weight (Maximum) - 2.5%
Neutral ammonium citrate soluble phosphates (P2O5) per cent by weight (Minimum) - 46%
Water soluble Phosphates (as P2O5) per cent by weight (Minimum) - 41%
Manufacturing Process:
Liquid ammonia is fed to the pre-neutraliser through dispersion distributor located at the bottom.
Phosphoric acid 54% P2O5 is injected into the pre-neutraliser from the top. Agitator with variable driven
motor is provided in the reactor for thorough mixing to yield slurry.
Chemical reaction take place in the diammonium phosphate fertilizer manufacturing pre-neutraliser
NH3+H3PO4 NH4H2PO4 (Mono ammonium phosphate)
NH4H2PO4 + NH3 (NH4)2HPO4 (Diammonium phosphate)
Figure 1: Process flowchart of DAP manufacturing
The resulting boiling slurry from pre-neutraliser is fed on the rotating bed of recycle material in a rotary
drum granulator and therefore it is ammoniated to 1.8 to 1.85 moleration by sparging liquid ammonia
through the formation of recent granules bed in the granulator. The moist granulated material discharged
from granulator is dried in a rotary dryer using hot-quenched gas to reduce the moisture content of the
material from about 3% to 1%. Cascade screen separated the dried material based on the size of granule
particles. The screened out particles are cooled by the rotary drum cooler the inclined rotation enhance huge
quantity of hot granules to contact with cool air blown counter currently from the outlet. The oversize
materials from the screens are pulverized and recycled to the granulator.
The grade of the DAP is adjusted by addition of urea and filler. Conveyor system transports the
Diammonium phosphate fertilizer to the bagging system. Scrubbing system traps the fumes obtained from
the granulator and pre-neutraliser and the dry scrubber cleans vent out air that carries the fertilizer particles
generated from pulverizers, screens and product cooler. Finally tail gas scrubber further scrubbers the air
before discharging it into the atmosphere.
Important parameters affecting quality:
The study of manufacturing process of DAP helped us understand the key parameters that affect the
specifications of the final product. The pre-neutralization process largely affects the final total nitrogen
percent and the phosphates percent. Also the granulator process decides the particle size of the final product.
So the pre-neutralization and granulation processes are studied in detail here to access the product quality.
Sampling Procedure:
Earlier method:
Usually, samples are taken the final product bagging area and tested for total nitrogen content and particle
size. But when an issue is detected, the damage is already done since a lot of out of specifications product
would have rolled out. To prevent this, we proposed a new sampling method.
New method:
In the new method, the samples are directly taken from the outputs of pre-neutralization and granulation
areas. Sixteen samples per day (one sample every hour) are taken from the output of the pre-neutralization
step to test total nitrogen content and total phosphates content. One sample per day is taken from the
granulator output to test the particle size. The test results are given as in a sheet shown below:
Figure 2: Daily analysis report for DAP shows N2 content, P2O5 content and particle size.
These samples adequately represent the samples from the final product bagging area. To prove this, a
difference of means hypothesis test is done. Samples are taken from the pre-neutralization area and also the
bagging area. The two sample t test is used to find if there is any difference in the means. The results are as
shown below:
Pre-Neutralization process monitoring:
Total Nitrogen Content:
The samples from the pre-neutralization output are tested for total N2 percent. The results are plotted on a
control chart and monitored. The limits of the control chart are calculated from the first 20 samples. The
data was plotted on the control chart and no anomalies were found. So these limits were fixed to monitor
the process.
The control chart for the N2 is shown below:
Change in supplier
of NH3
Xbar-S Chart of N2 Trial units
18.8
Sample Mean
UCL=18.594
18.4
__
X=18.146
18.0
LCL=17.697
1
11
16
21
26
31
36
41
46
Sample
Sample StDev
1.00
0.75
1
0.50
UCL=0.551
_
S=0.169
0.25
0.00
LCL=0
1
11
16
21
26
31
36
41
46
Sample
Figure 3: Control chart for N2 shows increasing trend after change in supplier of NH3
Out of control points were found in the 49th and 50th points. An increasing trend was also noticed in the XBar chart. This was reported to the production department immediately for analysis.
Cause of variation:
After analysis by the production department, it was identified that the variation was due to change in
supplier of Ammonia (NH3) that is used as an input to the process. To confirm this, NH3 from the old
supplier was used after discovery of OOC point and the process monitored. The process went back to
control. The control chart after the anomaly was corrected is shown below:
Figure 4: After changing supplier, the N2 content is in control
The supplier quality and the purchase departments were notified of the incident and action is being taken
with the new supplier. Due to the new sampling methodology, the defect was identified far sooner than
usual and further loss was avoided.
Total Phosphate Content:
The samples from the pre-neutralization output are tested for total P2O5 percent. The results are plotted on
a control chart and monitored. The limits of the control chart are calculated from the first 20 samples. The
data was plotted on the control chart and no anomalies were found. So these limits were fixed to monitor
the process.
The control chart for the P2O5 is shown below:
Xbar-S Chart of P2O5-Trial Units
41.7
UCL=41.655
Sample Mean
41.4
__
X=40.974
41.1
40.8
40.5
LCL=40.294
1
11
16
21
26
31
36
41
46
Sample
UCL=0.8361
Sample StDev
0.8
0.6
0.4
_
S=0.2560
0.2
0.0
LCL=0
1
11
16
21
26
31
36
41
46
Sample
Figure 5: Control chart for N2 shows increasing trend after change in supplier of NH3
No out of control points were discover during the duration of our project. The P2O5 content is in control.
Process capability of pre-neutralizer:
Sampling from the pre-neutralizer allowed us to calculate the process capability of that particular process
separately. This value will be added as a Key Performance Indicator for the whole manufacturing process.
Since two characteristics contribute to this process (N2 and P2O5 content), a sample is counted as defective
if one of them is out of specification. Then process capability is calculated by finding the Binomial
Capability as shown below:
Figure 6: Capability analysis of pre-neutralizer process shows that it is in 2.3 sigma level
The DPMO of the process is 170000 and the process operating at about 2.3 sigma level.
Granulation process monitoring:
The samples from the granulation output are tested for particle size. Minimum of 90% of the sample should
pass through a 4mm sieve and no more than 5% should pass through 1 mm sieve. If the samples dont pass
the test, they are marked as defective. Since the test is a Go/No Go test, an attribute chart (P chart) is used.
20 samples were collected initially and trial control limits were established (Figure 4).
The control chart is shown below:
Air flow rate
setting changed
Figure 7: Control chart for particle size shows increasing trend after changing setting of air flow rate
Out of control points were found in the 35th and 37th points. The process was deemed to be Out of control
and production stopped to analyze the cause. Upon analysis by the production team, it was found that the
granulator speed was changed to account for low temperature in the granulator. But since the team did not
have a model to find granulator speed for the particular temperature, they had to choose it arbitrarily. This
was causing defects in particle size.
Solution to the problem:
A design of experiment was proposed to determine the relationship between factors affecting the granulator
process and the output of that process. The parameter that affect the particle size are identified to be
granulator speed, granulator temperature and air flow rate. An experiment was designed to identify the
relationship between these parameters as shown below:
Figure 8: The design parameters of the experiment to determine optimal factor settings
The experiment needed a total of 44 runs. Data collection form were generated and sent to production
team for performing the experiment. The data collection forms are as shown below:
Figure 9: The data collection form given to the production team
The data collected from the production team was entered into Minitab for analysis. A linear model was fit
to the data to find the optimum setting needed for the desired output of 4 mm particle size. The results are
as shown below:
Figure 10: A linear model is fit to the data and the optimal settings are found
As shown in the results above, the optimum factor settings are found to be:
Granulator Speed: 200 rpm
Air Flow Rate: 2.31 m/s
Temperature: 583 degF
After this, regression analysis was done to generate the model equation. The equation generated by the
regression is given below:
Regression Equation
ParticleSize = 7.203 - 0.001344 Temperature - 0.3500 AirFlowRate - 0.008375 GSpeed
Further monitoring of particle size:
After implementing the corrective action, the particle size is monitored with the same control chart
without changing the control limits. The control limits will be changed in the future if a significant
improvement in the process is noticed. The control chart after implementing the corrective action is
shown below:
Figure 11: After applying the optimal settings, the particle size is in control
Process capability of granulator:
Sampling from the granulator allowed us to calculate the process capability of that particular process
separately. This value will be added as a Key Performance Indicator for the whole manufacturing process.
A sample is counted as defective if it does not conform to the specifications. Then process capability is
calculated by finding the Binomial Capability as shown below:
Figure 12: Capability analysis of granulator process shows that it is in 2.7 sigma level
Conclusion:
Results:
1. The sampling process is improved by sampling from the pre-neutralizer and granulator outputs
directly, instead of from the final product bagging area
2. The capability of the pre-neutralizer process is found to be 2.3 sigma
3. Design of experiments is done to find the optimal factor settings for the granulator
4. An equation to calculate the factor settings to achieve a particular particle size is generated
5. The capability of the granulation process is found to be 2.7 sigma
Suggestions for further improvement:
1. Gage R&R study of all instruments should be included in the daily analysis reports
2. Newer testing methods for finding nitrogen content should be used in place of current time
consuming test method
3. Automatic Samplers should be installed in the pre-neutralization and granulation output areas if
this sampling method proves to be more beneficial in the future
References:
1. http://processflowsheet.com/diammonium-phosphate-fertilizer-manufacturing/
