Sampling.
Bulk sampling of refining lots and subsequent sample preparation are complex, time-consuming and require meticulous attention. The aim of a good sampling protocol is to produce a set of homogenous assay samples that fairly represent the metal content of the lot. Most refiners have well equipped and well staffed sampling departments, but some locations are far from ideal.

Unless a representative sample is obtained, nothing can be done in either the assay procedure or the assay exchange to rectify the problem. In time and money invested, sampling is often looked upon as the poor relation compared to the assay work. However, in reality more money is lost through poor sampling than through poor assaying!

Here are a few key points to address when looking at a sampling protocol:
Ø      Does the location have the correct equipment to obtain a representative sample?
Ø      Prior to making the shipment, ask for the proposed sampling protocol. If you do not
       have expertise “in-house” to critically review the method, ask an experienced
       representative to review it for you.
Ø     The size of the bulk sample will depend on the value of the lot and on particle size
       distribution in the material.
Ø     The lot size for sampling and settlement should be based on the value of the material.
       It is sometimes better to break a single valuable lot into smaller lots to reduce reliance
       on a single sample.
Ø     A small deviation in a low-grade lot will have less financial impact than the same
       deviation in a high-grade lot of the same size.
Ø     Replication is more likely with a mechanical sampling system than with a manual
       sampling system.

  Some specific things to look out for are:
Ø     In metal lots, sampling is often when the lot or bulk sample is molten. Weigh the
       material into the melt, record any fluxing / melting agents that are added and record the
       metal and slag yield from the melt. The sample taken from the molten metal should be
       “shotted” or vacuum pin sampled. This will minimize segregation within the sample. The
       slag should be prepared and analyzed because metals can, and do, migrate to the slag.
Ø     If another metal is added to promote the melt, then the added metal should be assayed
       to ensure that payable or penalty elements are not being introduced into the final
       sample. If they are, a deduction must be made to the final assay.
Ø     In the circuit boards, sampling normally involves shredding and stream sampling. The
          resulting bulk sample can then be melted or burned prior to final preparation.
Ø     Prior to sampling sweeps or other fine materials such as baghouse dust, ensure that
       the lot is blended prior to sampling. Blending is important because material of different
       particle size will segregate during storage and transportation.
Ø     Excess grinding of the oversize fraction of a sample that does not pass a specified
       mesh screen can result in metal “smearing” on the grinding equipment, thus degrading
       the sample. In such cases it is better to retain the oversize and assay it to destruction or
       melt it (with the addition of another metal such as copper or lead) so that a homogenous
       sample of the oversize can be assayed.
Ø     Sludges should be dried and / or burned prior to sampling to remove water, oils and
       other volatile material.
Ø     Liquid samples are normally homogenized by stirring. Solids contained in the liquid
       must be evenly dispersed throughout the solution to be fairly represented in the final
       sample. Quantity is often governed by volume but it is also a good idea to obtain a
       weight because from this and the density volume can be calculated.

Assays and Splitting Limits.
The next step after obtaining a representative sample is to determine the metal content of both payable and penalty elements. How the metal content is determined can vary from lab to lab.  Laboratories will adapt their technique to the value of the lot and the level of accuracy required.

After completing the analysis, an assay exchange takes place. The exchange mechanics should be determined in advance and confirmed in the contract. This is also true of the splitting limits, the umpire procedure for when a third party result is required to make final settlement and choosing the umpire lab(s). How you set up and manage the assay exchange, splitting limits and umpire scenario can save or cost significant amounts of both time and money.

Some key points to guide you through this process are as follows:
Ø     Splitting limits are normally defined by commercial considerations rather than
       technical / laboratory capabilities.
Ø     Splitting limits are very important and are often negotiable.
Ø     Record and compare all assays. Trends can be detected in the performance of a given
       lab for a given element.
Ø     Exchange all silver results on a fully corrected basis.
Ø     If you do not have your own “in-house” assay facilities form a working relationship with
       a commercial laboratory. The lab should help you determine applicable lot sizes,
       splitting limits etc. The laboratory should welcome feedback on assay exchanges, for
       this will allow it to monitor the performance of its analytical procedures.

Going to Umpire.
Umpire assays are commonly used to settle assay exchanges that fall outside contractually agreed splitting limits. The contract between the two parties should fully define the umpire process. Although the rationale of an umpire assay is to moderate extremes, the umpire assay sometimes falls outside the two exchange assays rather than between them. Because of this, the method to determine the final settlement figure should be defined in the contact. Common examples are taking the middle of the three assays to govern or an average such as the mean of the two closest assays. The contract should also define how to determine who will be responsible for the umpire charges.

Some key issues to consider before you go to umpire are as follows:
Ø     A lab that has performed the exchange (party) analysis should not be used as the
       umpire laboratory.
Ø     In addition to fees, going to umpire can have an indirect cost if final settlement is
       delayed for an extended time. To minimize this cost, a provisional payment can be
       made based on the lowest exchanged assay. This should be defined in the contract.
Ø     Monitor the performance of the umpire lab(s) that you are using. A good umpire result
       can help two parties resolves a dispute amicably; a bad umpire result can annoy both
       sides. Plot your win / loss ratio over a number of data points. This can highlight
       problems with the umpire lab or with the assays you are using for exchange purposes.