Thursday, 11 October 2001 Media Statement
Lessons learned from diesel filter problems
Energy Minister Pete Hodgson today released a government report on the diesel-filter blocking problem that occurred
between March and May this year.
Extensive investigations into the problem found no clear cause for the filter blocking, which affected about 4 percent
of New Zealand's diesel engines.
"I asked both the New Zealand Refining Company and the Ministry of Economic Development to investigate the problem and
its likely causes," Mr Hodgson said. "Used filters and samples of diesel fuel were tested. Unfortunately no conclusive
results were obtained, although the tests indicated a possibility that a new winter additive may have contributed to the
blocking.
"The filter blocking problem is still subject to commercial and legal discussions amongst the companies involved. They
are still investigating the cause as part of this process and I understand this investigation is focused on the chemical
interaction between the additive and the diesel oils being produced by the refinery at the time. The results of this,
when known, should help clarify the cause."
"While it is frustrating that no clear cause has yet been found, some important lessons have been learned and action is
being taken to prevent this type of problem from recurring:
- the industry has ceased the use of the new additive and will be considering other options for additives for next
winter;
- the refinery has invited the Ministry of Economic Development to participate in a planned review of its quality
management systems, and;
- the Government is looking at regulatory changes to avoid similar problems happening in the future.
"The Government is publicly consulting on a review of the Petroleum Products Specifications Regulations and I anticipate
changes taking place from around mid-2002. Part of this review is the proposal for a new filterability test for all
diesel sold in New Zealand.
"These actions should give New Zealanders confidence that a similar problem with diesel fuel is highly unlikely to
recur."
ENDS
File: RN3/17/2
28 September 2001
Minister of Energy
REPORT OF THE BLOCKING OF FILTERS IN DIESEL-POWERED ENGINES - MAY-JUNE 2001
PURPOSE
1 This report has been prepared following the blocking of filters in diesel-powered vehicles in May-June of this year.
It describes:
- the use and supply of diesel in New Zealand;
- the regulation of diesel quality;
- the identification of the filter blocking problem;
- tests and analyses conducted to identify the extent and cause of filter blocking;
- the results of these tests and analyses; and
- the interpretation of the test results and analyses.
2 The report draws on:
- a detailed assessment of the blocking of filters undertaken on behalf of the New Zealand Refining Company (NZRC), by
Intertek Testing Services (Intertek);
- an analysis of blocked filters undertaken on behalf of the Energy Safety Service (ESS) of the Ministry of Economic
Development by Industrial Research Limited (IRL);
- tests undertaken on behalf of ESS by Inspectorate Singapore;
- statistical data on the number of insurance claims records; and
- a review of filter blocking claims received by Caltex NZ.
3 There may also have been reports of the filter-blocking incident undertaken by other parties that have not been
forwarded to the Ministry for its consideration and, therefore, could not be taken into account in preparing this
report.
EXECUTIVE SUMMARY
4 The results of tests and analyses undertaken suggest that no clear cause can be ascribed to the diesel filter-blocking
problem, although there are a number of test results suggesting the additive Dodiflow may be implicated in some
instances.
5 For a range of logistical and practical reasons, it did not prove possible to obtain a good number of used filters to
subject to testing and analysis. Intertek obtained 84 and IRL obtained 16 used filters. This compares to 16,502
reimbursement claims related to filter-blocking problems.
6 Diesel retailed met the requirements of the Petroleum Products Specifications Regulations 1998. These, however, do not
contain a filterability standard. It is proposed, as part of the review of the Regulations currently in progress, to
include a filterability test or other measure to avoid diesel filter-blocking, and to work with the industry to
determine the best approach.
7 An audit of the NZRC quality management system by Intertek concluded that there were adequate controls in place for
the processing, testing, and release of finished products. It concluded, therefore, that the filter-blocking problem
could not be directly attributed to the NZRC’s quality systems. As part of its continuous improvement process, NZRC is
reviewing its quality management systems and procedures. It has invited the Ministry to participate in this review. The
Ministry has accepted this invitation.
DIESEL USE IN NEW ZEALAND
8 Diesel is used in heavy vehicles, such as trucks and buses, and is becoming increasingly popular in cars and vans as
light diesel engine technology develops. Diesel is also widely used in the marine sector, powering fishing boats,
launches and yachts, and in agriculture for tractors, harvesters and other machinery. It powers railway locomotives and
all kinds of heavy machinery, such as cranes and earthmoving and construction equipment. Diesel is also used in
factories and buildings for stationary equipment such as generators and small boilers.
9 For the year 2000, New Zealand consumed around 2,400 million litres of diesel (and 2,900 million litres of petrol).
Eighty percent of diesel was used for transport, 6% for agricultural use and 6% in the fishing industry.
DIESEL SUPPLY
10 The Marsden Point Oil Refinery, operated by NZRC, supplies 90% of New Zealand’s diesel requirements and 65% of its
petrol requirements.
11 NZRC is owned as follows: BP 24%, Mobil 19%, Shell 17%, Emerald Investments 14%, Caltex 13%, Public 13%. The
ownership arrangements are somewhat unusual in the international refining business, where it is more common for oil
companies to own their own refineries.
12 The Marsden Point Oil Refinery has been operating since 1964. A major expansion was completed in 1986. It is
essentially a middle distillate refinery, configured primarily for the production of diesel and jet fuel, rather than
petrol.
13 The Refinery is a tolling operation; that is, it charges a toll on each litre of fuel produced. It processes crude
oil on behalf of the refinery users, BP, Mobil, Shell and Caltex. The refinery users individually purchase crude oil and
supply it to the Refinery to turn into refined products. The Refinery does not own the crude oil it receives or the
products it refines.
14 Diesel is produced at the Refinery to meet the mandatory quality requirements set out in the Petroleum Products
Specifications Regulations 1998 and the Refinery users’ own internal specifications for refined product (called the NZRC
and User Company Specifications).
15 The NZRC and User Company Specifications have been determined because the four major oil companies are common users
of the Refinery. They are in some cases more stringent than the Regulations and include additional specifications as
determined by customer requirements on a national or regional basis. Some companies also market additional specification
grades (for example, BP obtains lower sulphur diesel that it sells in the Christchurch region).
16 In order to ensure that product is produced to the quality required by the refinery users (and also that it complies
with the Regulations), NZRC undertakes quality control tests prior to releasing product for distribution. Tests
undertaken are the same as or similar to those specified in the Regulations for monitoring purposes.
17 Diesel is distributed from the Marsden Point Refinery by pipeline to the Wiri storage terminal in South Auckland, by
coastal tanker to storage terminals at New Zealand ports, and by road tanker to outlets throughout Northland and North
Auckland. NZRC owns the pipeline and the four major oil companies jointly own the coastal tanker network.
18 The balance of New Zealand’s diesel requirements not met by the Marsden Point Oil Refinery is imported directly,
predominantly to storage terminals at Wellington, Lyttelton and Mt Maunganui.
REGULATION OF DIESEL QUALITY
19 The Petroleum Products Specifications Regulations 1998 set out the standards that diesel (and petrol) retailed in New
Zealand must meet. Current diesel specifications impose controls on diesel’s physical and performance properties and on
a few individual constituents (for example diesel sulphur levels). The Regulations are currently being reviewed.
20 The purpose of the Regulations is to ensure that diesel (and petrol) available to New Zealand consumers is safe,
reliable and of good quality. The Regulations also include provisions for the protection of public health and the
environment. The Regulations were issued when diesel and petrol distribution was deregulated in 1988.
21 The Regulations sit within a wider consumer protection framework that includes the Fair Trading Act 1986 and the
Consumer Guarantees Act 1993. The Regulations provide buyers of petrol and diesel with an additional layer of protection
to ensure that fuel bought is fit for purpose while also allowing health, safety and environmental concerns to be
addressed.
22 Fuel quality is regulated because consumers are not in a position to determine the quality of fuel at the point of
purchase. Problems with poor quality fuel may not be experienced immediately. For the consumer, different brands of fuel
may have been bought over a period of time before a problem is identified, and it would be very difficult to prove the
cause of damage or any other problem.
COMPOSITION OF DIESEL
23 Both petrol and diesel are mixtures of many different hydrocarbons. Diesel is heavier and less volatile than petrol.
Due to its composition, diesel has a tendency to form wax crystals at low temperatures and, if these crystals are
allowed to grow, they can start to block fuel filters.
24 Fuel additives are available to overcome cold flow problems. Middle distillate flow improver (MDFI) additives
interfere with the formation of wax crystals, allowing adequate flow and filterability at low temperatures. These
additives vary in composition, but they are generally made up of polymer molecules of different sizes.
25 MDFI additives have been available worldwide since the 1960s. They are a cost effective means of producing diesel
that meets cold flow properties and of extending the proportion of distillates that can be used for diesels and other
mid-distillate gas-oils. NZRC has been using MDFI additives since 1997.
FILTER BLOCKAGES
26 In 2000, NZRC undertook a selection process to determine the ongoing supply of MDFI additive to be used at the
Marsden Point Oil Refinery. The selection process found in favour of Clariant’s Dodiflow 4032-2, (referred to in this
report as “Dodiflow”). This additive had not previously been used at the Refinery. Use of Dodiflow at the Refinery
commenced in March 2001.
27 On 23 May 2001, at a supply managers’ meeting, NZRC was informed of a diesel filter-blocking problem with NZRC fuel.
The problem was found to be affecting all four major companies.
28 On 25 May, it was concluded by the four major oil companies that the problem appeared to be associated with the use
of Dodiflow. A response was commenced aimed at mitigating the risks associated with use of diesel containing Dodiflow
and at providing a supply of diesel without the additive.
29 This response included:
- immediate shutdown of all marine diesel supply facilities until they were able to be supplied with diesel not
containing Dodiflow;
- press statements to warn diesel users that they may experience filter blocking difficulties;
- processes for quarantining diesel containing Dodiflow and supplying replacement diesel, in respect of marine users and
emergency services;
- industry help lines to provide advice to diesel users; and
- an industry commitment to replace affected fuel filters.
ANALYSIS OF EXTENT AND CAUSE OF BLOCKED FILTERS
30 In order to try to establish a cause of the increased fuel filter blocking being experienced by diesel users, various
analyses were instigated. Analyses of the extent of the filter-blocking was also undertaken.
31 Tests were undertaken by Intertek (on behalf of NZRC) and IRL (on behalf of the Energy Safety Service of the Ministry
of Economic Development) on actual blocked filters that they were able to obtain. Intertek was able to source 84 used
filters and IRL was able to source 16 used filters. It was not possible to obtain a larger number of blocked filters due
to a range of practical and logistical reasons, including the fact that many garages had already disposed of the filters
they had replaced and that reimbursement claimants were not required to surrender blocked filters as a condition of
payment.
32 Intertek subjected its used filters to a microscopy test, the purpose of which was to measure the degree of blockage
of the filter. It also subjected its used filters to chemical analyses to determine the cause of any blockage.
33 IRL subjected its used filters to visual examination to determine any signs of blinding material, and to flow rate,
molecular analysis and polymer tests to determine the cause of any blockage.
34 Intertek also undertook tests on laboratory filters of 0.8, 5, 8 and 12 microns, using diesel supplied from the
Marsden Point Oil Refinery that did not contain the Dodiflow additive but which they then dosed with typical amounts of
Dodiflow. Filtration, microscopy and chemical analysis tests were undertaken to determine any correlation of filter
blockage and Dodiflow.
35 Intertek also obtained 27 samples of typical diesel being retailed from service stations around New Zealand. These
samples were tested in the laboratory using filtration and chemical analysis tests to determine any correlation of
filter blockage and Dodiflow. Similarly, it undertook tests on NZRC diesel samples obtained from March 2001 known to
provide representative samples of diesel containing Dodiflow since its use began at the Marsden Point Oil Refinery.
36 Thirty-one samples of diesel obtained by the Energy Safety Service over the period December 2000 to June 2001 were
forwarded to Inspectorate Singapore and tested for filterability.
37 GAB Robins, the loss adjustment company appointed by the oil companies to co-ordinate reimbursement claims,
established and maintained a database of claims received.
38 Caltex NZ undertook a simple statistical analysis of blocked filter claims submitted to it. It correlated the area of
the country where the affected vehicle was used and the type of engine difficulty that was experienced to the known
diesel fleet and recorded concentrations of Dodiflow in each region.
39 Intertek undertook an audit of NZRC documentation and procedures relevant to the use of Dodiflow.
SUMMARY OF TEST RESULTS AND STATISTICAL ANALYSES
Intertek Tests of 84 Used Filters
40 Intertek’s report indicated that of the 84 used filters obtained, 45 were known to have come from North Island
locations, 5 from South Island locations and 34 were of unknown origin. Twenty-one of the total were known to have come
from fishing vessels and other boats, 28 from emergency service vehicles and light and heavy trucks, 6 from stationary
engines and 1 from a car. The uses of the remaining 28 filters could not be determined. Seventy-seven of the analysed
filters had been removed because of blockage concerns. An additional 7 filters were not considered blocked when in use
but were removed from engines and included in the testing as controls.
41 Filters were examined by electron microscope and classified as to degree of blockage based on this examination. A
chemical analysis of the material on the filters was performed to determine its composition. Particular attention was
given to determining the presence and concentration of polymers that are used in middle distillate flow improver
additives. Table 1 outlines the results of microscopy and chemical analyses of the filters.
Microscopy Analyses Chemical Analyses
No Polymer EVA PVA Total
100
g/cm2 >100
g/cm2
Mostly Blocked 1 9 6 1 17
Partly Blocked 3 23 1 2 29
Not Blocked 4 30 1 3 38
Total 8 62 8 6 84
Table 1: Intertek Analyses of Used Filters
Notes to table: includes 2 also with PVA
42 The polymers detected can be grouped into two general classes: ethylene vinyl acetate (EVA) and polyvinyl acetate
(PVA). EVA is commonly used in many middle distillate flow improver additives, including Dodiflow. PVA is used in some
diesel additives but was not a component of Dodiflow. Neither EVA nor PVA is found in diesel that does not contain
additives.
43 Table 1 shows that 70 of the 84 filters tested showed a presence of the polymer EVA that was similar in chemical
composition and size to the Dodiflow insoluble residue found in filtration testing (refer paragraph 57). Eight filters
had greater than 100 micrograms per square centimetre (g/cm2) which was taken as an indicator of a significant presence
of EVA, probably from Dodiflow.
44 A second type of polymer, PVA, was identified on a small number of the filters. PVA is not a component of middle
distillate flow improvers such as Dodiflow, and it is not clear what the source of the PVA was. The presence of PVA does
not seem to correlate with the degree of blockage observed in the filters analysed.
45 Thirty-eight of the filters showed no evidence of significant blockage, 29 were partly blocked and 17 were mostly
blocked. Of the 17 filters that were mostly blocked, 15 indicated the presence of EVA. Of the 8 filters with a
significant presence of EVA, 6 were identified by microscopy to be mostly blocked indicating a possible relationship
between the amount of EVA present and the amount of filter blocking observed.
46 Intertek report indicated that filters classified as dirty by visual examination generally had higher amounts of EVA,
as determined by chemical analysis.
47 Intertek found it difficult to draw conclusions based on examination of the used filters because of a lack of
information about the filters’ sources and histories and the relatively small numbers that were obtained. The presence
of other materials potentially blocking some filters was another factor that made it difficult to draw conclusions.
IRL tests of 16 Used Filters
48 IRL observed that of the 16 used filters it obtained some were so heavily burdened with general dirt and dust that
they may have passed their recommended operational life, or perhaps were taken from fuel systems that were unusually
dirty. Histories of the filters were not available.
49 Table 2 outlines the results of visual examination, flow rate, molecular analysis and level of polymer tests.
50 Filter flow rate tests were undertaken to assess the operability of the used filter. Filters are designed to operate
to standard even when 90% blocked with material. The flow rate test, therefore, looked at whether the flow rate was less
than 10% of a new filter.
51 Molecular weight analysis of material extracted from used filters was undertaken and compared with the molecular
weight range of Dodiflow. Molecular weight is a dimensionless measure of the mass of a molecule relative to a known
mass. Dodiflow contains molecules of many different sizes. Its most common molecular weight is about 10,000, but
Dodiflow also contains smaller amounts of molecules with molecular weights of up to 80,000.
52 Like Intertek, IRL also used chemical analyses to assess the level of polymer present on the filters.
Filter Blocked, possibly diesel related Filter blocked – dust, dirt, past recommended life Total
Visual Examination 12 4 16
Flow rate <10% of a new filter 1 3 4 Flow rate ≥10% of a new filter 11 1 12 Total 12 4 16 Molecular weight of organic material <20,000 2 0 2 Molecular weight ≥20,000 10 4 14 Total 12 4 16 Level of polymer ≤100g per cm2 8 3 11 Level of polymer
>100g per cm2 4 1 5
Total 12 4 16
Table 2: IRL Analyses of Used Filters
53 Table 2 shows that higher molecular weight material was found on 14 of the 16 used filters. Polymer analysis found
all of the 16 used filters had polymer present, with 5 having a significant presence (greater than 100 micrograms per
square centimetre). Flow rate analysis, however, did not indicate significant blockage (determined as being where flow
rate was less than 10% of a new filter) related to higher molecular weight material or polymer found on the filters.
54 IRL’s limited trials were unable to demonstrate conclusively that the polymer on the sample filters was responsible
for fuel blockage and that it was derived from Dodiflow. The quantities of extracted materials from the filters
suggested that, in some cases, the polymer may have caused reduced flow, but IRL was unable to confirm the blockage. It
considered this was possibly because the filters and fuel had subsequently been exposed to conditions that destroyed the
structure of the blocking gel. It noted that the thermal history of the submitted blocked filter samples may have
changed during transit and storage.
55 IRL suggested that the build-up of Dodiflow-related material may have been filter blinding rather than blockage.
Filter blinding refers to a reversible build-up of material. Filter blockage involves penetration of the filter core by
the blocking material. Blinding is considered reversible under favourable conditions, such as warmer temperatures or the
presence of diesel without Dodiflow. Filter blocking, on the other hand, is not reversible.
Intertek Laboratory Filtration Tests: A - Testing of NZRC-sourced diesel with Dodiflow Additive
56 Intertek tested the filterability of diesel containing Dodiflow at typical dose rates using laboratory filters of
0.8, 5, 8 and 12 microns. Intertek chose these filter sizes as automotive and marine diesel filters typically have pore
sizes of 4 to 10 microns. Some advanced fuel systems have filter pore sizes down to 2 microns.
Filter Size Time to filter
500mL (seconds) Volume filtered in
180 seconds (mL)
0.8 Failed 140
5 Failed 225
8 Failed 285
12 10 -
Table 3: Intertek Laboratory Filtration Test Results: Filterability of Dodiflow – dosed Diesel through various porosity
filters
57 By using a range of filter sizes, Intertek was able to gauge the size distribution of undissolved particulate matter
in the test diesel. In these tests, a defined amount of diesel (500 mL) was passed through a filter with a certain pore
size under vacuum. If the diesel passed through the filter in less than 180 seconds, the time elapsed was recorded. If
the diesel did not completely pass through the filter in 180 seconds, the sample was considered to have “failed”, and
the volume of diesel that did pass through the filter was recorded.
58 Table 3 outlines the results of these tests. The diesel failed to pass through the 8-micron and smaller filters, but
successfully passed through the 12 micron filter, which indicates that a significant proportion of the blocking
particles in the diesel were likely to be greater than 8 microns but less than 12 microns in size. Some of the blocking
material seems to have been retained on the 12-micron filter, though, as the sample that passed the 12 micron filter was
then passed completely through successively smaller pore size filters.
59 The filters used in these successive filtration tests were analysed to characterise the blocking material. Electron
microscopy was used to determine the size and shape of the blocking particles, while chemical analysis was used to
determine their composition. For all of these filters, it was established that the filter pores were completely covered
with material that was up to 5 microns in size and that this material was composed of an EVA polymer.
60 As a comparison, a filter that was used to test diesel lacking the Dodiflow additive was also analysed by microscopy
and chemical testing. It showed no evidence of blockage and no presence of EVA polymer.
Intertek Laboratory Filtration Tests: B - Testing of Retail Diesel Samples
61 Intertek tested the filterability of 27 diesel samples obtained from 5 June to 12 June 2001 from service stations
operated by BP, Caltex, Mobil, or Shell throughout New Zealand. Of these samples, 13 failed to pass through the 0.8
micron filter test described above. Of these 13, one sample also failed to pass through a 5 micron filter and one sample
was insufficient to test further. Six of the filters from among those that failed at 0.8 microns were analysed for
filter blockage composition by the methods outlined above.
62 By electron microscopy, it was determined that all six filters were completely covered with material that was up to 5
microns in size. Chemical analysis determined that the composition of the blocking material contained EVA polymer.
63 A second set of retail samples was obtained of diesel not likely to have been sourced from NZRC. Of these four
samples, all failed to pass through a 0.8 micron filter, and they were among the slowest of all the retail samples to
pass through a 5 micron filter. An analysis of one of the blocked 0.8 filters from this set of tests indicated that the
filter had a blockage similar to that formed by the first set of retail samples.
64 The results of these tests suggested that the retail samples tested, both from NZRC-supplied companies and from other
service stations, contained an insoluble residue that was very similar in composition to the Dodiflow insoluble residues
prepared in the laboratory.
Inspectorate Singapore Analysis of Filterability
65 Thirty-one samples of diesel collected since December 2000 were sent to Inspectorate Singapore, which tested them for
filterability using the IP387 method.
66 The samples sent to Singapore were retained samples left over from the routine fuel quality monitoring programme,
plus samples from June 2001. They were chosen to be a representative selection of samples collected since December 2000.
67 The scope of IP387 indicates this test is “… for the determination of the filter blocking tendency of distillate fuel
oils, where the end use demands an exceptional degree of cleanliness”. It is particularly applicable to installations
protecting gas turbines and other installations where there is a need for exceptional cleanliness and high throughput.
Its primary purpose is not as a filterability test for automotive diesel.
68 The result of the IP387 filterability test is an index number that indicates how filterable the diesel tested is. The
lower the index number, the more filterable is the diesel. Conversely, a higher index number indicates that the fuel is
less filterable and thus more likely to block filters.
69 Table 4 shows the filterability test results for the 31 samples. The samples are divided into three groups: before
the use of Dodiflow additive (first used at the Marsden Point Oil Refinery on 15 March 2001); after the problem of
filter-blocking was identified (post 25 May 2001); and a transition period between 15 March and 25 May. Samples with a
higher filterability test index number are more frequent after mid-March, with the three samples obtained after 25 May
all having a high filterability test index number. These results support the suggestion that diesel fuel was less
filterable after 25 May than before 15 March.
Number of Samples
Filterability Test
Index Number Before
15 March 2001 Transition After
25 May 2001
1.0-1.9 13 4 0
2.0-2.9 3 3 0
3.0-3.9 1 1 1
4.0-4.9 1 - 1
5.0-5.9 0 2 1
Table 4: Inspectorate Singapore Testing of Filterability of Diesel Samples Using IP387
Claims for Blocked Filters
70 As at 26 September 2001, GAB Robins reported that 16,502 claims were received through the oil industry call centre
relating to blocked filters. Of these initial claims, 11,292 have been submitted and approved for payment. A breakdown
of initial claims by vehicle type is shown in the figure below.
71 The graph indicates that of the claims, about 50% involved passenger cars, 23% were for light commercial vehicles and
9% related to trucks. Other claims related to marine vessels, tractors, diesel-using machines and other engines.
72 The number of claims compares with about 430,000 diesel vehicles and 1,300 fishing vessels powered by diesel in New
Zealand. In all, reimbursement claims represented less than 4% of the diesel vehicle fleet.
Caltex NZ Analysis of Claims it received
73 Caltex undertook an analysis of its claims for blocked filters. Although this analysis represents just a portion of
the total reimbursement claims lodged with GAB Robins, it is useful as an indication of the extent of the filter
blockage problem.
74 The number of claims in most cases seemed to be related to the size of the fleet in a given region, rather than the
concentration of the Dodiflow additive in the fuel. For example, Auckland had the lowest additive concentration but the
greatest number of claims. The greatest additive concentrations were present in diesel supplied to Dunedin and Bluff.
75 However, exceptions to this general rule existed. For example, Caltex considered that Nelson represented an
interesting situation in that it received one shipment of diesel with a very high concentration of Dodiflow. Although it
has a relatively small fleet, it had a similar number of claims to Wellington, which has a much larger fleet. This
observation seemed to support the link between the diesel containing Dodiflow and incidences of filter blockages.
76 On the other hand, Mt Maunganui, Wellington and Christchurch had a larger number of claims than expected, given they
were supplied with both imported diesel, believed not to contain Dodiflow, and diesel from the Marsden Point Oil
Refinery.
77 The number of claims directly caused by the diesel additive problem was likely to have been overstated. A breakdown
of the claims by engine sympton is given in Table 5. Symptoms involving a stopping/stalling element were definitely
related to fuel restriction (filter blockage) and were reported in 33.6% of cases. Other symptoms, which formed 66.4% of
claims, could have been caused by factors other than fuel filter blockage, such as mechanical malfunctions; for example,
plugged or restricted return; incorrect pump to injector timing; or air inlet restriction.
Nature of Symptom % of Total Definitely related to Filter Blockage Could be caused by Other Factors
Significant loss of power 41.8
Significant loss of power, hard starting 24.6
Significant loss of power, hard starting, complete stopping 21.3
Complete stopping 5.7
Significant loss of power, complete stopping 3.9
Hard starting, complete stopping 1.9
Hard starting, loss of power, stalling after starting 0.8
Total 100.0
Table 5: Category Breakdown of Reimbursement Claims
78 Other problems that can affect diesel filterability, such as contamination in the supply system, micro-organism
infestation in storage tanks, and lack of preventative vehicle maintenance may have contributed to the number of claims.
79 The actual number that can be attributed to the diesel additive problem cannot be accurately deduced. Due to
practicality and expediency, customers were not requested to send in clogged filters as a condition of compensation.
Consequently, fuel filters could not be tested to identify the cause of failure.
Intertek Audit of NZRC Documentation and Procedures
80 Intertek audited NZRC's documented quality system in relation to the production, testing and subsequent release of
diesel containing Dodiflow. The audit assessed:
- the adequacy of controls in place for process, testing and release of finished products;
- compliance with documented procedures; and
- the adequacy of the documented procedures.
81 The audit found that the controls in place regarding the processing, testing and release of finished products should
prevent inadvertent distribution of non-conforming products. There is zero tolerance set for any product that does not
meet the statutory requirements. At all times, NZRC supplied diesel within the requirements of the Petroleum Products
Specifications Regulations 1998. These do not require diesel, however, to meet a filter-blocking tendency test. Intertek
concluded that the quality system at NZRC was appropriate for using the Dodiflow additive and that the filter-blocking
problem could not, therefore, be directly attributable to the Refinery quality systems. However, Intertek did identify
areas where the Refinery could make some improvements in its overall operating procedures. These are being implemented
as part of NZRC’s continuous improvement model.
COMMENT ON TEST RESULTS AND ANALYSES
82 The tests and analyses undertaken indicated that:
a There is evidence that the Dodiflow additive was a factor in a number of filter blockages encountered during May and
June this year across New Zealand.
b EVA, a component of Dodiflow, was identified on 15 of the 17 mostly blocked filters analysed by Intertek, with 6 of
the filters having a significant presence of EVA.
c Of the 84 filters analysed by Intertek, 38 were classified not blocked; in other words, 49% of the tested filters,
excluding the 7 control filters, were not blocked.
d Polymer was found on all of the filters, and higher molecular weight material, possibly Dodiflow, was found on 14 of
the 16 used filters analysed by IRL. Flow rate analysis, however, did not indicate significant blockage related to the
higher molecular weight material or polymer found on the filters.
e Used filters classified as dirty (perhaps past recommended life) were observed by both Intertek and IRL to have higher
rates of polymer present. Filter blockage related to dirt, dust or perhaps the filter being past its recommended life
was also observed.
f Diesel containing Dodiflow mixed in the laboratory was shown in Intertek’s experiments to either block or restrict the
flow on 0.8, 5 and 8 micron filters.
g Samples of diesel from service stations analysed by Intertek also had insoluble residues, retained by a 0.8 micron
filter. These were very similar in composition to the insoluble residues retained on filters tested with diesel
containing Dodiflow that was prepared in the laboratory.
h Four samples of diesel not likely to have been sourced from NZRC tested for filterability also all failed the 0.8
micron filter test and were the slowest to pass through the 5 micron filter. It is not known whether a MDFI additive was
present in this diesel. The reason this diesel failed the filterability test appears to be insoluble particles similar
to those found in Dodiflow.
i Inspectorate Singapore test results indicated that diesel samples obtained after the use of Dodiflow were more likely
to indicate a tendency towards filter-blocking. This was particularly evident in the three samples taken after 25 May
2001.
j Less than 4% of the diesel vehicle fleet appears to have been affected by the filter-blocking problem. If Caltex’s
analyses typifies the problems experienced throughout New Zealand, it can be extrapolated that less than 2% of the
diesel vehicle fleet experienced filter-blocking problems. Of this amount, it appears likely that dirty filters were a
significant contributing factor to the filter-blocking problem in a number of cases.
k Diesel retailed met the requirements of the Petroleum Products Specifications Regulations 1998. The Regulations,
however, do not have a filter-blocking tendency test.
CONCLUSION
83 These results suggest that no clear cause can be ascribed to the diesel filter-blocking problem, although there are a
number of test results suggesting Dodiflow may be implicated in some instances.
84 The oil industry tests its batches of diesel (and petrol) prior to sale to ensure compliance with the Refinery User
Company Specifications and the Petroleum Products Specifications Regulations 1998. The blocking of diesel filters
experienced in May-June 2001, however, demonstrates that the specifications and testing are not always able to prevent
problems occurring.
85 The Petroleum Products Specifications Regulations 1998 aim to ensure that the diesel (and petrol) available for sale
meets certain specifications so that it is fit for purpose. Various properties are specified to provide quality product
that enables:
- good engine performance, reliability and service;
- starting and operating in all types of weather; and
- use that minimises risk to personal health, safety and the environment.
86 The Regulations are currently under review and a public discussion paper on proposed changes to the Regulations has
been released.
87 Two proposals to add to the Regulations that may reduce the risk of repetition of the unusual blocking of filters
experienced:
- a general clause requiring all fuel to be 'fit for purpose’; and
- a suitable diesel filterability test, to be complemented by introducing a filterability test to the routine Fuel
Quality Monitoring Programme.
88 The determination of a suitable test is still in question. None of the standard filterability tests is considered
fully reliable or able to address the particular problem New Zealand has just experienced. For example, IP387, used by
Inspectorate Singapore to test the filterability of diesel samples from December 2000 to June 2001, is a test
particularly applicable to installations where there is a need for exceptional cleanliness and high throughput. It was
not designed as a filterability test for automotive diesel. The Ministry has been advised that the tests currently
available are indiscriminate in what they will detect. This means that care has to be exercised in setting regulations
so that whatever measure is chosen will respond as intended.
89 It is recommended that a working group with industry be established to consider what filterability testing or other
options could best meet New Zealand's needs, given the lack of an obvious approach.
90 NZRC is in the process of reviewing its quality management systems and procedures and has invited the Ministry to
participate in this review. The Ministry has accepted this invitation.
RECOMMENDATIONS
91 It is recommended that you:
a note that no clear cause can be ascribed to the diesel filter blocking problem experienced May-June 2001, although
there are a number of test results suggesting Dodiflow may be implicated in some instances;
b note that the Petroleum Products Specifications Regulations 1998 are designed to ensure that an appropriate level of
fuel quality is available to consumers;
c note that the Regulations are currently under review and two additional measures are proposed that should reduce the
risk of repetition of the filter blocking problem:
i a general clause in the Regulations requiring all fuel to be ‘fit for purpose’; and
ii a suitable filterability test to be specified (assuming an appropriate test can be identified);
d note that there is not a readily available appropriate filterability test specification New Zealand could adopt;
e agree, in consultation with the Associate Minister of Energy, that the industry be invited to participate in a working
party to determine a suitable filterability test option or other measure to avoid diesel filter blocking;
f note that, as part of its continuous improvement process, NZRC is reviewing its quality management systems and
procedures and that the Ministry, at NZRC’s invitation, will participate in this review.
Graham Boxall Evelyn Cole
Manager Operations Manager
Energy Safety Service Natural Resources Policy
Recommendations agreed/not agreed
Hon Pete Hodgson
Minister of Energy