Feature Article

Palm Oil Mill OER and Total Oil Losses

Othman Walat* and Ng Say Bock **

T
INTRODUCTION areas’ of oil losses must be quantified and
subjected to regular independent checks.
he use of oil extraction rate (OER) The focus is on oil as the same can be drawn
as a tool to measure performance up for kernels.
has created a lot of problems for
the estates and mills. This is be- TOTAL OIL
cause OER on its own does not tell us where
the problem is. Whenever OER is low, the Barring security problems, the total oil com-
estates point their fingers at the mills and ing into the mill from the estate is the total
vice versa. In reality, OER is in ‘no man’s sum of what the mill recovers, which is OER
land’ and both parties do not want to be plus whatever the mill loses in the process.
held responsible for something that is not This can be represented by the equation:
within their control. Such ‘finger-pointing’
will continue as long as OER is looked at in Total oil = OER + oil losses
isolation without making any cross refer-
ence to oil losses. This article proposes the Rearranging the equation, we have:
incorporation of ‘oil losses’ into the OER
equation to complement the existing meth- Total oil – oil losses = OER
od of assessing estate and mill performanc-
This equation takes away the ambiguity
es. Once the fresh fruit bunches (FFB) are
surrounding OER and directs the responsi-
weighed in, the mill must be held respon-
bilities to the respective parties. This means
sible for all oil losses within the perimeters
that the estate will be responsible for the
of the mill. The estate, on the other hand,
total oil going into the mill while the mill
takes ownership of ‘total oil’, which is ‘OER
must provide an accurate account of what-
plus oil losses’. For the system to work, ‘all
ever oil they lose from reception of FFB
onwards. This, in effect, means that OER
* Innoprise Sdn Bhd, Sawit Kinabalu Sdn Bhd, Jalan is the joint responsibility of both the estate
Kelapa Sawit, Off km 4 Jalan Tuaran, 88300 Kota and mill. If both parties operate within their
Kinabalu, Sabah, Malaysia.
E-mail: [email protected] own satisfactory limits, it does not take a
** 61, 6th Floor, Jalan Melaka Raya 8, Taman Melaka genius to conclude that there will be a net
Raya, 75000 Melaka, Malaysia. increase in OER.

PALM OIL ENGINEERING BULLETIN NO. 108 11

 Feature Article

While OER can be accurately deter- loss is the missing link that must be quoted
mined and is subject to audit, the same can- in all discussions on OER.
not be said of oil losses. There is an urgent
need to review, expand and improve the Estate and Total Oil
monitoring of oil losses so as to capture the
‘total’ oil losses in the mill to complement The estate plays a critical role in ensur-
its use with OER. To be used effectively as ing high total oil in FFB at the time of har-
a management tool, these oil losses need to vesting. The adage ‘oil is made in the fields
be carried out or ‘audited’ by a third party and lost in the mills’, appropriately de-
at regular intervals. scribes such a situation. The key to high to-
tal oil starts right in the field with a correct
Low OER, Who is Responsible? harvesting policy.

The hypothetical OER is shown in Table 1. TOTAL OIL AND HARVESTING POLICY

In all cases, the OER is 18% and there are A harvesting policy is chosen to maximise
no clues as to why it is low. With the inclu- oil recovery in the field and it is influenced
sion of oil losses, we can infer in the first by two factors, namely: harvesting rounds,
case, that the estate is the cause. The sec- HR (days between harvest in the same field)
ond case points to milling problems while and minimum ripeness standards, MRS
in the last case; both the mill and estate are (minimum number of detached fruits per
jointly responsible for the low OER. It is bunch). For a given HR, selecting a MRS
obvious here that finger pointing between will fix the total oil in FFB of that harvest.
estate and mill will continue if we choose There is, however, a danger of lowering
to look at OER in isolation. Nobody wants MRS too much so that it borders bunches
to acknowledge responsibility for low OER without detached fruits. Table 2 (Woods et
and even if they know they are the cause, it al., 1984) shows the sharp difference in oil
is unlikely that they will announce it. The content in bunches with and without de-
inclusion of oil losses will take away the fin- tached fruits. They were harvested from
ger pointing and zero in on the cause. Oil one palm on the same occasion.

TABLE 1. HYPOTHETICAL OIL EXTRACTION RATE (OER) vs. MILL LOSSES

Estate Mill Management Responsibility of

Total oil - Oil losses = OER
(%) (%) (%)
Case 1: 20 - 2 = 18 Estate
Case 2: 23 - 5 = 18 Mill
Case 3: 21 - 3 = 18 Estate and mill

TABLE 2. OIL IN BUNCH WITH AND WITHOUT DETACHED FRUITS

Oil to bunch
Bunch ripeness
(%)
Black bunch, yellow mesocarp 13.6
Black bunch, red mesocarp 15.2
Red bunch, red mesocarp 20.8
Bunch with 31 loose fruits 22.3

12 PALM OIL ENGINEERING BULLETIN NO. 108

 Feature Article

Harvesting MILLS AND OIL EXTRACTION

It is common knowledge that in order to The mill’s role is to extract maximum oil
maximise oil recovery, the estate has to en- from the FFB with minimum oil losses. Un-
sure the following are adhered to: fortunately, process control monitoring in
• strictly maintaining the prescribed har- the mill is aimed at quantifying ‘controlla-
vesting rounds; ble’ losses, where adjustments can be made
• cutting only ripe bunches that is laid to reduce them. No emphasis is given to de-
down in the policy;
termine total oil losses. It is our opinion that
• making sure no ripe bunch is left be-
priority must now be given to quantify ‘all
hind; and
• picking up all loose fruits. oil losses’ within the perimeter of the mill.

It is of utmost importance that once a New Mills, Old Mills

policy is adopted; the estate’s role is to en-
sure its compliance to the letter. Most discussions on oil losses or OER
invariably end up on whether an old mill
Harvesting, the Practice gives a lower OER when compared to a
new mill. The relevant question should be:
In reality, there are a host of practical is- ‘how efficient are the mills in recovering oil
sues affecting harvesting as shown in Table 3. from the fruits it receive?’ This is tabulated
in Table 4.
The Impact of Labour Shortage/Poor Su-
pervision If we were to base our assessment on
OER, we will conclude that Mill A is an in-
Low total oil in FFB is a direct result of
efficient mill. In actual fact, Mill A is more
selecting the wrong harvesting policy, cut-
efficient in recovering oil from the fruits as
ting black or unripe bunches and leaving
behind loose fruits. Missing out on ripe it has a higher oil milling efficiency. Age of
bunches result in rotten bunches in the next the mill does not come into the picture and
round. The impact on total oil is summa- is of no consequence in oil recovery. An in-
rised below and it is important to consider efficient mill, whether old or new, will show
them from an economic standpoint in deci- up as having high oil losses and has a low
sion-making on issues related to harvesting oil milling efficiency. The real need is for the
practices. ‘oil losses’ to be accurately quantified be-
fore making the comparison.
Unripe : a harvest with 5% of each
bunches category of unripe bunches Process Control Measurements
without loose fruits (see
Harvesting, the Practice The starting points for measuring total
section) will depress the total by-products produced are shown in Table 5.
oil to bunch by close to 1%.
Poor loose : for a 20 kg bunch, every 10 It is imperative to use accurate flow me-
fruits loose fruits left uncollected ters and weighing equipment to quantify
collection lowers the total oil to bunch by by-products being produced. The lack of
about 0.2%. emphasis and the absence of suitable meas-
Trash : bringing in 5% non-oil bearing urement equipment coupled with the inher-
tissues artificially decreases to- ent difficulty in measuring have led some
tal oil to bunch by 1%. mills to estimate these measurements.

PALM OIL ENGINEERING BULLETIN NO. 108 13

 Feature Article

TABLE 3. ISSUES AFFECTING HARVESTING

Cause Effect
Insufficient harvesters Extended harvesting rounds
- Departure from harvesting policy
- Rotten bunches
Insufficient loose fruit pickers Poor loose fruits collection
Poor supervision Black and unripe bunches
High unharvested bunches
High uncollected loose fruits
High trash in FFB

TABLE 4. OIL EXTRACTION RATE (OER) vs. MILL EFFICIENCY

Mill A Mill B
(i) Oil losses 1.93 2.50
(ii) OER 20.80 21.20
(iii) Total oil [(i)+(ii)] 22.73 23.70
(iv) Oil milling efficiency [(ii)/(iii) x 100] 92 % 89 %

Bulk Densities cannot be overemphasised here because in

the final analysis, ‘the test results are only as
An example is given here to highlight good as the samples taken’.
the inherent difficulty in quantifying fibres.
A 30 t FFB hr-1 mill operating for 16 hr will OIL LOSSES IN THE MILL
produce 62 t of fibres. Installing a weighing
equipment to measure this is not practi- An example of process control monitoring
cal as it is equivalent to weighing 34 lorry of oil losses in a mill is listed in Table 6.
loads of fibres daily (calculated from a bulk
-3
density of 130 kg m and a lorry volume of The bulk of the oil losses are in empty
3
14 m ). In such a case, the total fibre is de- bunches, press cake fibre and sludge. A crit-
termined by weighing the total nuts pro- ical and often neglected area where losses
duced for the day and then using the fibre can be high is in the recovery of unstripped
to nut ratio test to quantify the total fibres bunches. Poorly sterilised bunches can run
produced. It is clear that nuts have to be ac- up as high as 40% and they have to be man-
curately weighed. The environment and the ually recovered for re-sterilisation. Losses
type of weighing machines used play a part from poorly recovered unstripped bunches
in its effectiveness in such measurements. have been shown to be as high as 0.5% on
FFB.
Sampling and Testing
Oil Losses not Measured
Most mills have comprehensive docu-
ments on how these should and must be The areas of oil losses not measured in
done. This however, will not be elaborated process control are:
here. The importance of adhering to proper • dumping of fruits on hopper top;
sampling procedures, techniques for sub- • loose fruits dropping between hopper
sampling and testing of the by-products grates;

14 PALM OIL ENGINEERING BULLETIN NO. 108

 Feature Article

• fruits spilled and crushed in the mar- Nonetheless, they have to be identified and
shalling yard; measured and included in to cover all areas
• drainage of process tanks; of losses within the mill.
• oil overflows from tanks;
• desander discharges; The Need for Quantifying Total Oil Losses
• tank washings;
• purifier and centrifuge flushing; and However small the oil losses are, they
• oil spillages. have to be determined because they have
already been accounted for when the FFB
The bulk of these ‘losses’ are gener- were weighed in. There could be inherent
ally recovered and recycled. Some of these difficulties in measuring some of these by-
losses are small and may appear inconse- products. If they are proven to be burden-
quential, as they do not happen all the time. some, then constants should be used to

TABLE 5. STARTING POINTS FOR MEASURING TOTAL BY-PRODUCTS

By-product What is measured? How it is measured?

Empty bunches A sample of fresh fruit bunch is The EB recovered from the fresh fruit
(EB) weighed for sterilisation bunch are weighed after threshing

Unstripped Visual count of the number of Calculating difference between theoreti-

bunches (USB) USB produced cal and physical recovery

Nuts Total nuts produced Basculator/flow estimation

Fibres None Derived from fibre to nuts ratio
Sludge Total sludge produced Basculator/flow estimation

Effluent Total effluent produced Basculator/flow estimation

TABLE 6. OIL LOSSES AND MILLING EFFICIENCY

Source of oil loss Oil loss on FFB

Steriliser condensate 0.16
Empty bunches 0.56
Fruit loss in empty bunches 0.03
Unstripped bunches 0.02
Press cake fibre 0.55
Nuts 0.06
Sludge/decanter 0.46
Washings/spillages 0.09
Oil losses 1.93%
Oil extraction rate 20.80%
Total oil (1.93 +20.80) 22.73%
Oil milling efficiency (20.80/ 92%
22.73 x 100)

Note: FFB – fresh fruit bunch.

PALM OIL ENGINEERING BULLETIN NO. 108 15

 Feature Article

‘signify’ that there are also other areas of oil CONCLUSION

losses that are not included in the process
control oil loss assessment. Interpreting OER on its own without men-
tioning its associated oil losses can be er-
Practical Constraints in Quantifying Oil roneous and can lead to inaccurate conclu-
Losses sions. With the inclusion of total oil losses,
the responsibilities of both estates and mills
Having stressed the importance of deter-
are clearly defined putting an end to finger
mining total oil loss in a mill, there are also
pointing when OER is low. It is necessary
practical limitations in getting such meas-
urements. How does one go about quan- for such important assessment to be carried
tifying oil losses at the hopper top? In the out by an independent party. There is also
process of pushing the fruits into the hop- an urgent need, however, to review and im-
per, the tractor runs over some fruits crush- prove the existing methods of measuring
ing them and leaving the oil on top of con- oil losses in the mill. In addition to this, ‘all
crete surface and on the tractor tyres. Even oil losses’ that occur within the confines of
if one can find a way to quantify them, how the mill from the time the FFB is weighed
accurate can such measurements be? Then in must be accounted for. It maybe a mam-
again, this measurement cannot be carried moth task but it is one, which has to be car-
out on a daily basis without disrupting mill ried out if one is serious about improving
processing. estate and milling efficiencies.

Dumping of FFB on the hopper top must

REFERENCE
be avoided at all cost. It is prudent to invest
in more cages or hoppers to avoid such oil
WOOD, B J; SAID ISMAIL; LOONG, S G;
losses and end up with quality problems.
If it must be dumped, then a series of te- and CHEN, S C (1984). A preliminary re-
dious tests are required where the FFB are port on a long-term study of the effect of
weighed before and after dumping on the oil palm harvesting strategy on product re-
hopper top. The difference in weight can covery, including a comparison before and
be multiplied against fixed oil in fruits con- after weevil pollination. Proc. of the Sympo-
stant and adjusted for moisture loss. Such sium on the Impact of Pollinating Weevils on
readings should be applied each time the the Malaysian Palm Oil Industry. PORIM and
FFB is dumped on the hopper top. MOPGC, Kuala Lumpur. p. 214.

16 PALM OIL ENGINEERING BULLETIN NO. 108