LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
Chapter 8
Quality Control
and
Analytical Methods
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 1of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
Chapter 8 Contents
8.1 INTRODUCTION
8.2 WHY IS CONTROL OF QUALITY IMPORTANT?
8.2.1 Bottled Gas
8.2.2 Industrial Gas (burner and furnace requirements)
8.3 WHAT ARE THE IMPORTANT SPECIFICATION ITEMS?
8.3.1 Vapour pressure control
8.3.2 Icing troubles
8.3.3 Toxicity and odour control
8.3.4 Corrosion
8.4 WHAT SPECIFICATIONS EXIST?
8.5 A GUIDE TO SPECIFICATION THINKING
8.5.1 Composition
8.5.2 Vapour Pressure
8.5.3 Sulphur Compounds and Corrosivity
8.5.4 Residues
8.5.5 Water
8.6 TEST PROCEDURES AND ANALYSERS
8.6.1 The Importance of Sampling
8.6.2 Sample Containers
8.6.3 Analytical Procedures
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 2of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
Chapter 8
Quality Control and Analytical Methods
8.1 INTRODUCTION
We propose to begin this chapter with a discussion on the reasons for
imposing some form of quality control on marketed LPG. These reasons will
be apparent from an outline of the main outlets for LPG and of features which
are required to ensure trouble-free operations. Since in practice the
optimization of LPG quality is brought about within the framework of
specifications, we continue with a critical review of certain international LPG
specifications, i.e. of those quality items which are selected for control in the
form of concentration limits, e.g. a maximum sulphur content, a minimum
propane concentration, etc. Because this critique shows up certain anomalies
and differences between specifications which are accepted by producers and
users alike at the present time.
Specifications are meaningless without prescribed test methods which permit
the particular quality feature to be determined accurately at the specified level
of concentration. Therefore, the chapter ends with a brief description of
analytical procedures which, in practical experience, have proved to be the
most useful for this particular purpose.
8.2 WHY IS CONTROL OF QUALITY IMPORTANT?
As noted in Chapter 6, LPG rarely consists of pure C 3 or C4 hydrocarbons
alone. Rather, is it a refined light portion of crude oil or natural gas
condensate, specially tailored at the refinery to suit a particular field of
applications. It follows that commercial LPG is a mixture of hydrocarbons with
possibly some impurities left over from the refining process or introduced
between the points of manufacture and consumption. Quality control
therefore becomes a matter of matching properties to applications, or
applications to properties, and minimizing impurities without unduly
increasing the product price.
The nature of established uses for LPG gives a clue to some of the properties
which are either necessary or desirable for this product to possess. These
outlets are discussed at length in later chapters. For the present, the main
ones can be summarized along with desirable LPG properties as follows:
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 3of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
8.2.1 Bottled Gas
Maximum vapour pressure commensurate with the container design
stress
Natural vaporization at year-round temperatures
No corrosion of container walls
No blockage or fouling of pressure regulator
No toxic impurities in the gas
Minimum non-vaporizable residues
Protection against icing on expansion of the vapour
Detectable smell
8.2.2 Industrial Gas (burner and furnace requirements)
No corrosive action of liquid, or formation of deposits or residues in
storage tanks, transfer lines, vaporizer, pressure regulators
No corrosion or fouling of burner tips
Controlled combustion characteristics (calorific value, Wobbe Index,
flame speed)
No catalyst fouling impurities (endothermic gas generators)
No toxic impurities (agriculture and food industries)
No freezing of separated water preventing liquid offtake or blocking
gas ports, etc.
This outline of applications indicates that some form of control over many
aspects of LPG quality is necessary if trouble-free operations are to be
guaranteed.
Control over product quality is exerted in two ways. One is by the product
distributor or the user setting a specification which suits the equipment or
process into which the LPG is fed. Such a specification usually names a
potentially harmful impurity and sets a limit on the concentration which can be
tolerated. The other way is by the LPG producer setting his own
manufacturing specification to match the user's requirements and by
continuously testing quality by means of stream analyzers or laboratory
checks. Both of these quality control methods are usually adopted in practice.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 4of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
8.3 WHAT ARE THE IMPORTANT SPECIFICATION
ITEMS?
The more important items for quality control are evident from the foregoing
outline of applications. They can be grouped together and summarized as
follows: vapour pressure; icing propensity; odour and toxicity; corrosiveness;
non-vaporizable residues and certain other trace impurities.
8.3.1 Vapour pressure control
Is readily understood to be a matter of balancing the content of the main
C3/C4 hydrocarbon components against lower and higher boiling homologues,
e.g. of specifying limits as follows:
Composition
Maximum content: ethane, ethylene, pentanes
Minimum content: propane, propylene, butanes, butenes
8.3.2 Icing troubles
Icing troubles are due to the presence or accumulation in LPG of water
which may separate and freeze if the liquid temperature is low enough,
or combine with the hydrocarbons or vapour to form relatively high
melting gas hydrates. If the pressure is high enough, the presence of
water also leads to rusting of non-stainless equipment and rust, in turn,
can add to the residues problem. Therefore, a limit is usually placed on
the content of water in one or all of three possible forms.
Water
Free or entrained: 'none visible'
Dissolved: maximum content
Another way of inhibiting icing problems is by the addition to LPG, more
particularly to propane, of an anti-icing agent such as methanol. This lowers
the temperature at which water can separate from the ternary hydrocarbon-
water-methanol system. Thus, recommendations are often made regarding
the concentration of methanol to be incorporated in LPG, particularly during
the winter season.
Methanol content:
minimum content (to prevent under-dosing)
maximum content (to avoid contamination problems)
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 5of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
8.3.3 Toxicity and odour control
Toxicity and odour control is a matter of minimizing the content of hydrogen
sulphide but allowing, in commercial grades, a very small proportion of
certain volatile sulphur compounds such as ethyl mercaptans or dimethyl
sulphide to impart a detectable odour. (For the highest purity grades, e.g.
aerosol propellant LPG, the smell may have to be minimized by eliminating
these sulphur compounds as well as any other odoriferous compounds such
as unsaturated hydrocarbons.
Odour and toxicity
H2S:
maximum content (toxicity)
Mercaptans (volatile sulphur compounds):
minimum content (detectable smell)
maximum content (catalyst poisoning, no odour)
8.3.4 Corrosion
Corrosion of metal equipment is of course controlled by limiting the corrosivity
of the product. Two methods of control are adopted in practice. One is to
specify the degree of corrosion which is allowable under standardized test
conditions, namely a slight discoloration of a copper strip immersed in liquid
LPG for a specified time and at a prescribed temperature (e.g. 1 hr, 37.8C).
Copper is chosen because it is the most sensitive of metals likely to be
encountered by LPG in ordinary process equipment. The other method is to
limit the concentration of known corrosive agents which could be found in
LPG. These agents are mainly H 2S and elemental sulphur. Additionally, in
very critical circumstances, control over corrosivity may have to be extended
to cover trace impurities which could conceivably cause corrosion over
extended periods. Such impurities or contaminants could include carbonyl
sulphide, water, oxygen, ammonia, caustic alkali and dissolved chlorides.
Corrosivity
As to non-volatile residues, these can comprise a miscellany of compounds
ranging from incompletely separated refinery fractions such as pentanes and
hexanes to the higher boiling contaminants such as lubricating oil, as well as
certain sticky deposits from rubber hoses which were described in Chapter 3.
They also include other forms of involatile material such as rust, dirt and
elemental sulphur. As has been mentioned, the lighter ends (C 5's, C6's) and
also elemental sulphur are specified separately for other reasons.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 6of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
8.4 WHAT SPECIFICATIONS EXIST?
Such specifications as exist at the present time derive from many
authorities, e.g. national (U.K., U.S.A., European, etc.), the LPG
supplier (natural gas producer or oil company), industry (i.e. association
of LPG interests) and user. They all relate to the type of LPG (propane
or butane) which is used, to the type of application in question (e.g.
domestic, industrial, chemical or automotive) and to the product grade
(commercial or high purity).
Table 8.1 Specifications for Propones
(a) NGPA, commercial
(b) NGPA, high purity
(c) British Standard Specification 4250
(d) Deutsche Normen 51622
(e) French Ministry of Industry & Agriculture CSR15A
Item included in specification
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 7of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
A comparison of these specifications in terms of the items specified and the
limits imposed are shown in Appendix 4, Tables 4(a) and 4(b) for propane
and butane grades, respectively. Here, it is sufficient to highlight the fact that
specifications differ in detail from country to country (see Tables 8.1 and 8.2).
The only items which are common to most specifications are vapour
pressure, mercaptan-type sulphur compounds and free water. Some
authorities also control corrosivity by inserting the copper strip test with a
permissible discoloration limit; others adopt the indirect method of imposing
limits on the corrosive agents, normally H2S.
Table 8.2 Specifications for Butanes
(a) NGPA, commercial
(c) British Standard Specification 4250
(d) Deutsche Normen 51622
(e) French Ministry of Industry & Agriculture
Pass on *Doctor Test* required to indicate freedom from H 2S and
mercaptans
Item included in specification
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 8of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
The impurities most commonly omitted from user and some suppliers'
specifications are many of those included in DIN 51622, i.e. elemental
sulphur, corrosivity and oily residues and elements such as fluorine, chlorine,
caustic soda, ammonia, oxygen, nitrogen, etc. However, it must be
emphasized that a very restrictive specification serves no useful purpose
unless the process concerned is known to be especially sensitive to certain
aspects of fuel quality.
8.5 A GUIDE TO SPECIFICATION THINKING
8.5.1 Composition
LPG is primarily required to be a source of fuel gas. The bulk composition of
the LPG will determine the combustion characteristics of the gas. If the liquid
LPG is to be vaporized naturally as in the case of bottled gas, the LPG
should have a high content of either C 3 or C4 hydrocarbons or else the
composition of the gas will be markedly different according to whether the
bottle is near-full or empty. For industrial purposes, however, the liquid is
always vaporized by external heating so that the characteristics of the gas
remain constant as long as the liquid composition does not vary. In the latter
case therefore, it is unnecessary to specify the exact composition of LPG
propane or butane in terms of C 3/C4 contents. In both cases though,
pentanes and hexanes should be limited to minimize residues. For propanes,
the restriction could be 'nil C 5's and above', and for butanes 'nil C 6's and
above' with a 2% maximum on C5's and higher. The unsaturates content
might be qualified in respect of critical applications such as, possibly, catalytic
reforming processes and automotive fuel outlets.
It must also be mentioned here that the so-called 'volatility' specification test
which measures the temperature at which 95% by volume of the propane or
butane is evaporated is, in fact, an index of composition. In the particular
case of commercial propane, the limit usually placed on this temperature is
38.3C. If this limit is to be met, the content of 'C 4' and above' cannot amount
to much more than 2.5 vol. %. Unless therefore the propane is particularly
required to have a minimum content of 95% C 3's approximately, it will be
found convenient to omit this test from specifications.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 9of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
8.5.2 Vapour Pressure
A maximum vapour pressure is essential to ensure that LPG meets the
requirement of pressure vessel design codes which apply in the countries
concerned with handling the product. A minimum value may also be required
to provide for a satisfactory gas flow rate by natural vaporization at low
ambient temperatures. In very cold climates, a minimum vapour pressure
may also be required to guard against vacuum conditions developing inside
storage tanks containing a high content of normal (n) butane. This in turn
would entail ensuring a minimum content of propane in butanes.
8.5.3 Sulphur Compounds and Corrosivity
The common practice of restricting corrosivity as judged by the copper strip
test is a convenient means of limiting by the same token the concentration of
the two types of corrosive sulphur compounds, namely H 2S and elemental
sulphur. These may therefore be omitted from quality specifications unless
specifically demanded for some aspects of process control. When required,
levels recommended range up to 2.5 ppm weight (H 2S) and 2 ppm (elemental
sulphur), respectively. This leaves volatiliable sulphur compounds like
mercaptans to be specified for odour control, a level of 25 ppm is adequate
for this purpose and a reasonable maximum is 50 ppm, though some existing
specifications allow up to 180 ppm weight total sulphur compounds.
8.5.4 Residues
A convenient limit to place on oily residues as judged by an evaporation
procedure is 500 ppm by volume. This can be tolerated without undue
nuisance to the user of residues building up in liquid vaporizers, though in
some critical instances, the 50 ppm weight limit already advised in certain
specifications may be found appropriate.
8.5.5 Water
The final specification item is water and two forms should be mentioned, free
or entrained, and dissolved. The former should be related to a bottoms LPG
sample from tankage and specified as 'non visible'. Dissolved water has to be
determined by analysis. For full protection against the most critical icing
conditions, the limits should be 10 ppm weight (propanes) and 20 ppm
(butanes), respectively. If the LPG is not refrigerated or dried to provide for
this low water content, a recommendation should be included for the addition
of compensative methanol (usually 0.05 vol. % min). Restrictions on trace
impurities should be imposed only if they are known to cause a particular
application problem.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 10of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
8.6 TEST PROCEDURES AND ANALYSERS
It is pointless to include a limit in specifications which cannot be measured
accurately by either party involved in a dispute about LPG quality. Therefore,
it is necessary to discuss existing analytical procedures for all the
specification items previously mentioned. Where alternative methods exist, it
is also necessary to pinpoint the most satisfactory technique as far as
sensitivity, accuracy and repeatability are concerned.
8.6.1 The Importance of Sampling
Quality control starts at the point of production. Any departure of quality from
an optimum or specified level is made evident by onstream analysers with the
necessary corrective measures taken either manually or by means of an
automatic signal from the analyser back to process. Unfortunately, automatic
on-stream analysers are expensive, sometimes unreliable, and require
frequent inspection and maintenance. Therefore, they are not viewed as
favourably in practice as perhaps they should be. Recourse is made instead
to intermittent sampling and laboratory analysis. This means that there is
always a lag between actual and measured LPG stream composition.
The origin of the sample is very important, and the point and moment of
sampling must be chosen in relation to the purpose for which the analysis is
being carried out.
Another important point is size of sample. It has pointed out that reducing the
liquid level of LPG inside a sample container increases the vapour volume.
This in turn increases the mass of a volatile impurity (such as ethyl
mercaptans or water) in the vapour phase and correspondingly decreases its
concentration in the liquid phase. Thus analysis of successive liquid samples
for volatile components can be expected to show diminishing concentrations.
The best solution is, therefore, to use a sample container whose volume is
large in proportion to the quantity required for analysis and not to further
divide this sample into separate containers. Alternatively, one should take a
fresh sample from the bulk source (refinery stream or tank) for each
successive analysis, i.e. to work on a 'one test, one sample' basis.
Some further confusion can arise from the selection of either liquid or vapour
phase LPG for analysis and the formulation of results, on a weight, mole or
volume per cent basis.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 11of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
Most analytical methods for LPG are based on gaseous samples, although
the bulk of industrial users are more concerned about the quality of the liquid
LPG which is supplied to them. Therefore it is essential to carry out every
analysis on a totally vaporized, liquid sample. This ensures that no high
boiling components of the liquid are left unaccounted for. It also prevents any
bias due to preferential concentration of volatile impurities in vapour standing
above the liquid, as in bottled (domestic) LPG. It is also suggested that
results are most conveniently expressed on a weight basis, i.e. as weight per
cent (wt %) or parts per million (ppm).
8.6.2 Sample Containers
Two main types of sample containers are used. One is the stainless steel
double-valve, cylindrical 'bomb' which ranges in size from 5 ml to 4.5L water
capacity. The other is the free-standing domestic-type bottle, with single or
twin gas valves, taking samples ranging in size from 3 to 50 kg. For replicate
determinations of any single specification item 200 ml 'bombs' are suitable.
For multiple determinations of a full range of items, a 25 kg bottle provides
the requisite quantity without undue loss of components from liquid into the
vapour phase between analyses.
Connections between LPG source and sample container should be made in
stainless steel or aluminium tube. Nylon tubing is satisfactory in small
diameters but flexible rubber or plastic (PVC/polythene) connections are not
recommended on account of the leaching effect of the LPG on soluble
components (i.e. plasticizer). Containers should be thoroughly cleaned and
purged between samples.
'Bombs' are sometimes fitted with outage or 'ullage' tubes to ensure at least a
20% volume free vapour space to allow for liquid expansion with temperature
rise. In sampling refrigerated propane, the ullage space should be increased
to 30% for increased safety. If no outage tube is fitted, the maximum fill can
be calculated and provided for by weighing the bomb before and after filling.
The sampling procedure is to hold the bomb upright with the outage tube
suspended from the top valve. Introduce a small quantity of liquid LPG
through the bottom valve and crack open the top one so as to cool the bomb
by liquid vaporization. Continue filling until liquid escapes from the top. Close
both valves disconnect from sample line, open top valve, until only gas (no
liquid) escapes.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 12of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
If the vaporized LPG is to be sampled on industrial plants, water-cooled steel
or ceramic suction probes are recommended for insertion into the flow lines.
Gas can be withdrawn into water-filled glass sample tubes, aspirator bottles
or metal gas holders. Plastic balloons are suitable only for short-term storage
of gas samples and if 'chalked', may eliminate any acidic impurities before
analysis.
8.6.3 Analytical Procedures
Hydrocarbon Composition
The most common method used is gas chromatography. A routine volatility
check method which reports the temperature at which 95% by volume of the
LPG is evaporated also indicates composition, as has already been
explained.
In gas chromatography, a liquid LPG sample is vaporized in a stream of inert
gas and passed through a gas-liquid partition column, maintained at a given
temperature, which separates the LPG vapour into its component
hydrocarbons. The components are detected in the effluent gas by means of
a thermal conductivity or flame ionization detector, or similar detection device,
using pure hydrocarbons as a check on relative retention times. In this way a
series of 'peaks' are obtained in a chromatogram, each peak representing a
pure component whose concentration is proportional to the area under the
peak.
Squalene can also be used successfully when only saturates are present,
and propylene carbonate if unsaturates have to be resolved. Thus, LPG can
readily be analysed to 0.01 mol % of the following components; methane,
ethane, ethylene, propane, propylene, n-butane, isobutane, butene-1,
isobutylene, cis-butene-2, trans-butene-2, isopentane, n-pentane, hexanes. A
typical LPG chromatogram is shown in Figure 8.1.
Laboratory gas chromatographic practices have also been applied to process
control of LPG stream composition. In the case of a propane-butane splitter,
this means resetting temperature and reflux ratio.
The NGPA volatility test (ASTM D 1837) is used with a maximum of
38.3C for the 95% evaporation temperature in the case of propanes and
+2.2C for butanes. Correlation work shows that the 38.3C temperature is
met if the content of 'C4' and higher is not more than 2.5 %; if the butanes
content in propane were 10% for example (as is allowed in B.S. 4250), the
evaporation temperature would rise to 23.9C.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 13of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
The limit of +2.2C is met by all butanes containing not more than 2% of
'pentanes and higher', a composition which is typical of most commercial
butanes.
Figure 8.1 Gas chromatogram of LPG (British Gas Corporation)
Vapour Pressure, Specific Gravity and Heating Value
The compositional analysis of an LPG provides a convenient basis for
calculating several properties, namely vapour pressure, specific gravity,
motor octane number and heating value. The method uses the fact that any
of these properties can be summed from the partial contributions of the
various hydrocarbon components, e.g. in the case of the mean vapour
pressure at 37.8C (v), allowing for main components only
v = av' + bv" + cv'" .
Where a, b and c are the molar fractions of propane, n-butane and isobutane
derived from the gas chromatographic analysis; v', v" and v'" are the
respective vapour pressures of the pure components at the reference
temperature (see Table 8.3). The other properties are calculated in similar
fashion.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 14of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
Table 8.3 Factors for Calculation of Physical Properties
In strict terms, summing of partial values applies only to ideal gases. For real
gases such as propane and the butanes, account should be taken of their
compressibility. Thus the true heating value for example, H' is obtained by
dividing the apparent value derived from the compositional analysis (H) by
the mean compressibility factor applying to the mixture (Z), which can be
calculated from the component values thus:
H' = H / Z
Apart from calculation methods, vapour pressure can be measured directly
using a thermostatted bomb, suitably 'ullaged' to allow for liquid volume
expansion between sampling and measurement temperatures, connected to
a pressure gauge.
Specific gravity of the liquid phase LPG can be determined using a pressure
hydrometer.
Hydrogen Sulphide
Two types of method for assessing H 2S concentration can be used. One is an
inferential method based on the staining of lead acetate paper. The other is
quantitative and is carried out by converting the H 2S to a mineral sulphide
and estimating the same either calorimetrically or titrimetrically.
For regular monitoring of LPG production, the authors recommend using a
vaporized liquid sample and testing the gas by a stain procedure selected
from those described according to the detection limit requiredwith
occasional use of the colorimetric methylene blue method for checking the
actual H2S content.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 15of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
Total Sulphur
The so-called 'total' sulphur content of LPG is only a measure of volatile
sulphur compounds (H2S, light mercaptans and low boiling sulphides, etc.)
which pass into the gas phase when the liquid test sample is evaporated at
about 80C into a test burner. This technique takes no account of elemental
sulphur or relatively involatile forms of sulphur compounds which may be
present in the liquid LPG.
An alternative method preferred by the authors involves electrometric titration
of an alkaline solution of the mercaptans using silver nitrate solution, with a
glass/silver sulphide electrode system to indicate the end point. The
procedure enables hydrogen mercaptan (H 2S) to be determined separately
from other mercaptans (methyl, ethyl, etc.) which are determined collectively.
Elemental Sulphur and Involatile Sulphur Compounds
There are four main methods, the mercury test, the Sommer test, a
colorimetric method and a combustion technique. All four depend on the
preliminary extraction of the sulphur by bubbling a known weight of the LPG
through toluene. The toluene solution is heated to remove any remaining
LPG or volatile compounds.
Oily Residues
A number of test methods are available but selection depends on the type of
residue that is to be determined.
As already mentioned, gas chromatographic analysis is available to measure
the amount of low boiling C 5/C6 residues present in LPG. If the technique
could be extended to include the higher boiling residues in the C 6 to C30+
range, which are sometimes encountered in practice, it would be highly
recommendable. Unfortunately, the an is not yet at this stage.
In the industry, residues are usually taken to refer to 'oily residues', i.e.
evaporated residues after removal of C 5/C6 fractions leaving mainly
lubricating oil. One test method entails the evaporation of about one liter of
the LPG at room temperature and the further heating of the residue on a
steam bath for five minutes in order to remove unwanted pentanes and
hexanes. The final residue is weighed. The drawback of this test is that the
residue will contain all non-evaporated material boiling above heptane, i.e.
gas oil as well as lube oil in addition to any elemental sulphur, dirt and rust
present in the sample.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 16of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
A more reliable measure of the lube oil content can be obtained by 'infra-red
spectroscopy'.
In all these residue methods, a word of caution needs to be added. The use
of plastic or rubber connecting tube should be avoided. What might pass as a
mineral oil residue often turns out to be a plasticizing compound such as
phthalates or ricinoleates leached out by the LPG from plastic tubing used in
the sampling or test equipment. Before reporting, therefore, the identity of the
residues should be checked by infra-red examination.
Water
A number of different methods are available. They include the following:
1. Chemical reaction with Karl Fischer reagent
2. Colour change with an indicator such as cobalt bromide
3. Dew point method
4. Gas chromatography
5. Instrumental techniques giving an instantaneous reading on a
calibrated scale
6. 'Valve freeze' method
The Karl Fischer method is reliable and accurate down to 20 ppm of water
in LPG. A known volume of LPG vapour is passed through anhydrous
methanol. The water absorbed is titrated with Karl Fischer reagent, i.e. a
mixture of iodine, sulphur dioxide and pyridine. The main reaction, in which
the pyridine does not take part, is:
H2O + I2 + SO2 + CH3OH 2HI + CH3HSO4
Gas chromatographic techniques have already been outlined for
determining hydrocarbons. Water can be separated from the hydrocarbons
on a column containing polyethylene glycol on a polystyrene bead support.
Trace Impurities
Ammonia: A number of different methods have been proposed, namely
titrimetric, colorimetric and indicator.
Chlorine: Total chlorine comprising both inorganic and organic forms is
determined by reaction with sodium diphenyl dissolved in toluene. The
method is sensitive down to the 1 ppm level.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 17of 18
LPG TECHNOLOGY Chapter 8 Quality Control & Analytical Methods
________________________________________________________________________________
Fluorine: The basis of the method is again the removal of fluorides and
fluorine compounds from the sample by bubbling the gas through a toluene
solution of sodium diphenyl.
Carbonyl Sulphide: This can be determined down to 0.1 vppm by gas
chromatography using a Porapak-Q column.
Oxygen and Nitrogen: These gases can be analysed at the 1 ppm level by
gas chromatography using 60-80 mesh 13 molecular sieve and a thermal
conductivity detector.
________________________________________________________________________________
ETS (Rev. 1) Nov. 2006 Page 18of 18
