HUMAC

Analytical Tanzania Laboratory Audits

Laboratory
(Mwanza)

HUMAC (Alex Stewart) Analytical Laboratory (Mwanza)

Audit: [Overall Perspective]

Confidential

Audit Report
20th – 21st August 2004

Document Number: 20040910

Reported to: Vaughan Chamberlain [AngloGold Ashanti]

Dereck du Preez

AssayTech
12 Glenanda Village
Le Roux Street
Glenanda.
Johannesburg
R.S.A.
2091

Tel: +27 11 4322560

Cell: +27 82 7746610
Email: [email protected]

Issue 1 Page 1 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

1.0 Introduction:

The Evaluation and Mining Geology Department of AngloGold Ashanti requested that the Analytical
Laboratories currently used, and those anticipated for use in Tanzania be audited.
It is generally accepted as sound Laboratory Management, and widely recommended that Analytical
Laboratories each year, be independently audited.
It is further acknowledged, that an internal audit of the facilities, be conducted within the same
year.
Regular Laboratory auditing thus contributes to entrenchment of sound analytical practices, thereby
assisting in maintaining the sought analytical standards, which enhance the Laboratories credibility
with customers.

2.0 Audit Procedure:

The Overall Perspective Audit provides a comprehensive overall assessment of the Laboratory and
the facilities that are available. It further views the more pertinent aspects regarding the various
standard requirements and analytical quality sought by AngloGold Ashanti. In addition, all
equipment and the condition thereof are documented. Brief summaries of the observed analytical
practices are recorded, together with other pertinent information.
The audit includes pictorial support.

3.0 Audit Results:

Results of the audit are comprehensively documented. Comments on the general performance are
provided, together with observed health and safety practices, training facilities, analytical
procedures, including remarks on the Laboratories overall anticipated capability and quality of
performance. Included in the report are results obtained on the provided Rocklabs Certificated
Reference Materials, these were analysed during the audit, with selected observation of certain
analytical practices.

4.0 Conclusions:

A comprehensive overview is reported. Suggestions, which may prove beneficial, are offered.
The report is intended to provide an Overall Perspective of the viewed facility, its equipment and
general routine analytical practices, including the anticipated ability, in providing the sought
standards.

Issue 1 Page 2 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

1.0 Overall Perspective:

HUMAC Laboratories (T) Ltd (Alex Stewart) are situated in an ”industrial area” approximately 10 to
15 Kms outside Mwanza.
The Laboratories are house in 3-4 buildings.
The main service is Fire Assay for gold exploration samples. (However, it is reported that the
Laboratory is receiving some Geita Mine grade control samples). In addition ppb and ppm gold
determinations by Aqua-Regia/MIBK extraction with F.A.A.S. measurement are offered. A small
number of base metal analyses are also conducted by acid decomposition and F.A.A.S.

The Laboratory is supported by Alex Stewart Laboratories in Ireland, and it is reported that these
Laboratories offer a wider range of facilities including ICP (OES & MS).
The HUMAC Laboratory does not operate with a Laboratory Information Management System.
(LIMS) so all work is manually managed using Microsoft Word and Microsoft Excel support.
All sample aliquot measurements are therefore conducted by the measurement of actual masses
rather than “catch weights”. In addition all other analytical data such as F.A.A.S measurements etc
are manually recorded.
Procedural Documentation is available in the Laboratory Manager’s Office.

In house Laboratory control is managed with “Geostats Standards” and the Laboratory tends
toward a negative bias. Copies of the Laboratory statistical control provided for perusal.

Typical of Laboratories visited in Tanzania no “lockout” facilities appeared on any equipment.

Health and Safety, and Environmental Policies were neither posted nor available. Access to
emergency medical assistance would have to be organized via the Laboratory Manager or Shift
Supervisors.
Blood lead monitoring is conducted on a 6 monthly basis, and the records retained by the
Laboratory Manager.

Training is “On The Job” (OTJ), and task based.

The facility does to an extent suggest “makeshift”, and would benefit from some organization.
However, some areas were “reasonable”, but the view is that the “Sample Preparation Laboratory”
would benefit by some immediate attention. This area appears old and dusty. Adherences to PPE
requirements appear haphazard.

No cyanide is used on the site, and acid neutralization is conducted using sodium hydroxide, this
is allowed to trickle to a “storm water drain”.

Waste from the Fire Assay procedure is accumulated (slag, fusion crucibles, cupels etc) and it is
reported that this material is buried on site in concrete.

Only “dry sample preparation” is conducted.

Initial sample storage is in containers for pulp samples, and a large garage type shed for rejects,
however, this is for a limited period, thereafter samples are returned to the client or disposed with
the clients consent, (“bulldozing and covered over on the HUMAC site”).

The facility can manage 600 samples per day and operates on Mains Power with own Diesel
Power Generator as backup.

Issue 1 Page 3 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

2.0 Sample Receiving/Sorting/Drying:

All samples are received and sorted on wooden tables under a roofed area. Received samples
arrive packed in Calico, and Plastic Bags Photo 1 & 2.
Samples are sorted into numeric sequence, and referenced to the delivery documents. Once
confirmed, the consignment data is entered into the computer.
In the case of the Calico bagged samples, these are dried in the bag. Those in plastic are
transferred to dishes (made on site) Photo 3. Dishes are washed between uses, Photo 4.

Photo 1 Photo 2

Photo 3 Photo 4

Sample drying is conducted in a container, which is kerosene fired Photo 5. The temperature of
the facility is well maintained and this was confirmed on a number of occasions during the audit
Photo 6.

Photo 5 Photo 6

Issue 1 Page 4 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

3.0 Sample Preparation/ General Equipment:

The room size is reasonable. However, it does appear rather “makeshift”. Photo 7 & 8 (Photo 8
notice foot attire).

Photo 7 Photo 8

The facility is equipped with 5 crushers (1XLabtechnics Essa, 1X Rhino 1X TM Engineering

Cone Crusher and 2X Marcy Crushers). Photo 9, 10,11 & 12 give a general impression of the
type of crushing facilities that are available.

Photo 9 Photo 10

Photo 11 Photo 12

Issue 1 Page 5 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

The majority of Pulverizers are Labtechnics LM2’s and these appear as though they have been
“used”. Photo 13 & 14. In addition to the LM2 Pulverizers, 3 X Horizontal Disc Pulverizers are
also available Photo 15.

Photo 13 Photo 14

Photo 15

Photo 16 & 17 shows samples in the process of being milled.

Photo 16 Photo 17

In Photo 17 two samples and two bowls are at the workbench. The situation was intriguing, and
warranted observation. As was anticipated, the operator was not sure of which sample had
come from which packet? This supports the recommended practice of only one sample at a time
at the workbench. In this way situations such as the one that was in this instance encountered,
would be avoided.

Sample splitting is conducted by “Jones Riffle Splitter” Photo 18. However, close observation will
show that the ‘third pan” is not used, the sample is placed in the “Chute type pan” and poured
through the splitter vanes. This is poor practice and was pointed out.

Issue 1 Page 6 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

Photo 18

No other types of splitting facilities were seen to be available.

Cleaning protocols, as with all Laboratories viewed in Mwanza, is achieved by the use of “Lake
Victoria Granite”; air hoses are typically used for cleanup. At HUMAC, construction of a “Dust
Box” has been attempted. This is a small building behind the preparation workbench into which
the dedicated extraction fan at the workbench blows. Photo 17.
Photo 18 & 19 shows the interior of the “Dust Box” behind the building into which the fan
extracts the dust emanating from the workbench.

Photo 17 Photo18 Photo 19

Fire extinguishers are available in the Laboratory, and it is reported that there are 9 in total.
Photo 20 & 21 shows both the powder and the CO2 variety. Both have been recently serviced.

Photo 20 Photo 21

Acetylene Cylinders are caged and locked; however the facility did appear “crowded”.
Photo 22 & 23. However, the cage is located on a solid wall, and the cylinder in use is fitted with
a “ flash back arrestor”. No records of routine inspections were available. The introduction of
such a register, was recommended

Issue 1 Page 7 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

Photo 22 Photo 23

4.0 Quality Control (Crushers & Pulverizers):

Screen tests are 1 in 200? It was recommended that these be conducted more frequently. In the
case of Crusher products it is reported that these are between, 58% to 95% < 2mm. In the case
of LM2 Pulverizers 86% to 97% < 75 micron.
No computer-generated records are available, at present all data is recorded in a workbook. But
statistical data is being prepared such that notice board display can be implemented.

5.0 Fire Assay:

The area is relatively large. Cleanliness is typical of a busy Laboratory Photo 24. The facilities
include 1X Propane Fired Furnace (recently rebricked) Photo 25. 1X Diesel Fired Furnace
currently out of service (waiting to be rebricked) Photo 26. There are also 3X Cupellation
Furnaces one of which is waiting rebricking

Photo 24 Photo 25 Photo 26

The Cupellation Furnaces are electrically fired. The condition of the firebricks appears sound,
and it would seem that they are timeously rebricked.
Premixed flux is sourced from South Africa (no litharge quality certificate). Collectors used are
silver, and this is added by dispensor.
De-Slagging appeared sound, Photo 27. The anvil was kept free of slag and the buttons were of
adequate mass Photo 28
Typically copper maps assist in tray control, however two batches during the visit were found
with identical patterns? Photo 29.

Issue 1 Page 8 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

Photo 27 Photo 28 Photo 29

The pouring technique is the single pot pour (it is advisable that the pot just be placed at a 450 –
500 angle to avoid the tendency to “toss out” any remaining lead droplets, and not “thrown over”
as depicted) Photo 30. Furnace Fusion Crucible loading are 6 pots simultaneously Photo 31.
Mould condition is shown in Photo 32

Photo 30 Photo 31 Photo 32

Notable in the HUMAC Fire Assay procedure, very few crucibles tended to boil over Photo 33.
A further point of note, are the storage of crucibles. These are stored according to the gold
values obtained on previous fusions, pots where the sample values were >5 g/t are discarded.
Photo 34 shows stored pots in racks with values assigned.

Photo 33 Photo 34

Issue 1 Page 9 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

An interesting point is that all samples after mass measurement are roasted at 5500C prior to
fluxing. Photo 35 samples being mass measured prior to the roasting step.

Photo 35

A concerning feature of the operation, was that lead fume escaped directly into the atmosphere.
Photo 36.
Photo 37 shows the accumulation of Furnace Room debris waiting eventual on site burial in
concrete

Photo 36 Photo 37

In some areas PPE is worn, and in other areas not, although the Heat Resistant Apron was
worn together with a Face Shield and Gloves in the Furnace Room, some employees were seen
to be wearing “street shoes”?

6.0 General Laboratory:

The area is of acceptable size and reasonably tidy, the water source is that of distilled Photo 38,
the quality of which is monitored. Prill decomposition takes place after transfer to test tubes and
dissolution is achieved by the use of temperature controlled Heating Blocks (5X50) Photo 39.
Hot Plates are sound and reagent labeling acceptable. Photo 40
A concerning point regarding Prill transfer from cupels to test tubes, was that the rack of test
tubes was placed behind the cupels, and the Prills transported to the test tube with no reference
into which test tube the Prill should be placed (other than “eye and memory”). This task should
be addressed, and either the test tubes be “racked” upside down, and only inverted right side up
upon placing of the bead into the tube, or alternatively, “finger indicate the test tube into which
the bead should be placed”.

Issue 1 Page 10 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

Photo 38 Photo 39 Photo 40

It was also noticed that dilution to volume was “mechanical” it would be suggested that
dispensers gravimetrically calibrated be used to dispense required volumes. This is a more
controlled and accurate procedure.

7.0 Instrumentation:

The only instrumentation is Flame Atomic Absorption Spectrophotometers (F.A.A.S.) of which

there are two, a Varian AA50, and a Varian Spectraa 5. Photo 40 & 41. Both Instruments appear
sound. However, the Instruments are not on a maintenance schedule, but are managed in
house. It is advised that a qualified technician regularly service these facilities.
Registers for the logging of instrument calibrations are available, however no second primary
standard solutions for calibration confirmation are used. It was suggested that the use of these
be considered.
No other instrumentation is available in the Laboratory.
Sample presentation to the F.A.A.S. is manually conducted.
Calibration Standard Solutions are prepared from 99.99% gold plate sourced from Rand
Refinery.

Photo 40 Photo 41

8.0 Balance Room:

The Balance Room was “busy”. Balances are not on a service contract; it is advised that this be
considered. Balances, as do all other instrumentation, require regular servicing and should be
maintained in a temperature-controlled environment.

Issue 1 Page 11 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

Balance Registers to record calibrations are in place, together with Standard Mass Pieces, but
no Certificates of Authenticity of the Mass Pieces are available. Photo 42 & 43 show the balance
room.
In addition room doors in which Air Conditioners were installed were left to stand ajar. They
must be closed at all times to maintain room temperature and avoid potential draft.

Photo 42 Photo 43

The facility has 1X Micro Balance, 2X Analytical Balance and 2X Toppan Balances.

9.0 Warehouse and Sample Storage:

Chemical and bulk storage is managed in containers.

Storage of sample rejects is in a garage type store Photo 44.Some bags had weathered. Pulp
storage is in a secure store Photo 45.

Photo 44 Photo 45

Sample retrieval is managed manually.

10.0 Analytical Quality Control:

“Geostats” standard materials are placed on each tray of samples as routine, together with
blanks and client standards. HUMAC results are accordingly recorded and statistically tabulated
Each year a suite of “Geostats” materials are received from “Geostats”. These are analysed and
the values returned to “Geostats” for compilation and evaluation, a “Certificate of Achievement”
is awarded. Copies of the Certificate appear in the “hand out brochure”
HUMAC have partaken in this exercise for some time.
Generally, the Laboratory tends toward a negative bias. However with more controlled
procedures in certain areas, it is possible that improvement in the bias can be achieved.

Issue 1 Page 12 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

During the “Overall Perspective Audit” Rocklab Standard Reference Materials were handed to
the Laboratory Manager. These were analysed with selective procedural observation by the
Auditor.
Tabled are the results together with the recommended values.

International Reference Material Recommended value HUMAC Laboratory [Mwanza]

Rocklabs Cert. Material SJ10 2.643g/t Au 2.60g/t Au
Rocklabs Cert. Material OXH29 1.298g/t Au 1.25g/t Au
Rocklabs Cert. Material SJ10 2.643g/t Au 2.62g/t Au
Rocklabs Cert. Material OXF28 0.802g/t Au 0.79g/t Au
Rocklabs Cert. Material SK11 4.823g/t Au 4.95g/t Au

Perusal of the above data indicates some samples tend to a low bias one however, tends
toward high.
The Laboratory Manager manages Quality Control. Presently all statistical evaluations are within
the computer system, but at present there is no Notice Board display thereof.

It is reported that certain staff have an understanding of statistical controls.

11.0 Reporting:

There is no LIMS and all reporting is therefore manually conducted and takes the form of a
spreadsheet, unless otherwise requested. This will be customized as per requirements. Billing is
managed from the HUMAC site.

12.0 Staffing/Training:

The total number of staff is 32, of which 2 are Graduates and 7 Trained Technicians. In addition
there are 23 unskilled trained members of staff. The Laboratory works two shifts.

Training is On The Job (OTJ) and therefore task based, Planned Task Observations (PTO’s)
assist in supporting operations. No one possesses a First Aid Certificate, and it may prove
prudent to consider this. Safety statistics are reported to be good.

13.0 Health & Safety:

Health and Safety appeared haphazard; it did not appear as though this was enforced. It
seemed as though staff wore the PPE if they so wished; however in certain areas they appeared
to ensure the wearing of certain items. It would be suggested that a more firm approach to
Health and Safety be adopted.

Issue 1 Page 13 of 14
 HUMAC
Analytical Tanzania Laboratory Audits
Laboratory
(Mwanza)

In addition it is suggested that safety signs and their appropriate display in the strategic areas be
addressed.
A further area for consideration would be that of lead fume to atmosphere.

14.0 Conclusion:

The Sample Preparation Laboratory would substantially benefit by some urgent attention.
Some interesting procedures were observed. Notably roasting of the sample prior to Fluxing,
although not “new”, so few Laboratories conduct this step.
The interesting part of this procedure will be when high sulphide materials are encountered, and
such samples endure over a period of time. The SO2 generated will render the atmosphere in
the “roasting area” intolerable.

Generally it is believed that discounting the Sample Preparation area, that the Laboratory has
the ability to provide sound assay. The point that should be concentrated upon is that of prill
management, final volume dilution prior to instrument evaluation, and F.A.A.S. calibration (2nd
Primary Standards).
It must also be pointed out that sound procedures must at every workstation be undertaken with
purpose, such dedication will ultimately provide rewards of standard.

It would also be suggested that the entire laboratory be “walked” to ensure that all appropriate
Safety Signs are posted in the respective work areas. That all staff stringently practice sound
PPE requirements thus ensuring their Health and Safety.

In addition it is believed that fume and dust extraction should also be addressed, thus ensuring a
responsible approach to the environment.

Dereck du Preez.
(AssayTech)
Johannesburg (10th September 2004)

Issue 1 Page 14 of 14