EDITORIAL — Fire-assay checks should be prescribed — Lessons learned the hard way

In early 1994, the salting gang at the Busang ounce factory in Kalimantan realized that the coarse placer gold they were adding to drill-core samples caused “reproducibility problems” between duplicate samples when traditional fire-assay methods were used. After comparing results from “coarse gold-bearing” (that is, salted) samples sent to two different laboratories, geologist Michael de Guzman wrote a memo to his boss, John Felderhof, suggesting that on-site metallurgist Jerry Alo be given the task of finding a way to get reproducible results.

De Guzman’s concern was a valid one, as many salting scams, including the infamous Karpa Springs project in Australia, were exposed when someone took a close look at the reproducibility problems caused by coarse gold in samples. Having spent time Down Under, De Guzman was no doubt familiar with the story, which involved his former jungle workmate, Michael Novotny.

Noranda planned to pay up to $16 million for the high-grade project, which Novotny and his partner, Laurie Whitehouse, had optioned from a group of prospectors.

The deal was interrupted mid-stream when a seasoned Noranda geologist noted a problem with coarse, visible gold that caused “nugget effect” during assaying. Duly warned, Noranda’s Australian team performed some random checks and found that the results did not mesh with those previously reported, a phenomenon that always occurs when samples are doctored with coarse gold. The major refused to complete the transaction when the rejects mysteriously disappeared. The three prospectors who made the “discovery” — Robert and Dean Ireland and Clark Easterday — were eventually convicted of fraud.

Many other salting scams involved erratic results and reproducibility problems, including Trans-Terre, New Cinch, Dessir, Golden Rule and Bre-X, to name a few. Core went missing in the Tapin Copper salting scam and, in the latest salting scandal involving Mongolia Gold Resources, rejects mysteriously disappeared.

The salters at Busang knew they had good reason to worry that fire-assay results would expose their phony deposit. After some experimentation, they decided to abandon the 50-gram fire assay in favor of cyanide-leach analysis, which used 750 grams of material. The immediate result was a dramatic improvement in the reproducibility of analyses, which masked a tell-tale sign of salting. While the stated goal was to make the results more representative of what was in the core, the real intent was to make them less representative.

About the same time, Felderhof and de Guzman decided to send most of the full core for assaying, eschewing the standard practice of core splitting.

Felderhof told analysts that he was using large samples and cyanide extraction to avoid the problems experienced at the nearby Kelian gold mine, where a bitter lawsuit had been triggered by under-reported grades. The argument that Bre-X was merely following Kelian’s methodology served the needs of the salters, as it meant split core would not be available for later examination and, possibly, re-assaying.

Trying to find out what happened at Kelian was difficult because court records were sealed. But we have since learned that during the exploration stage Kelian did not use large samples, nor did it use cyanide extraction.

And core was split in the normal manner.

The problems came about because in the Kelian ore some of the gold came in coarse particles: about 45% of it was recoverable by gravity circuit. The operator submitted 10-gram samples, which were analyzed by atomic-absorption spectrometry following aqua regia digestion and back-extraction using an organic solvent. Duplicate samples of low-grade material yielded closely comparable results; in high-grade mineralization, duplicate results varied over a wide range because a single particle of coarse gold could bump the grade of a 10-gram sample sky-high. The operator, rather than investigating why this was happening, calculated an overall grade from statistical curves.

As the old saying goes, “garbage in, garbage out.”

The point of all this is that it would be a serious mistake for the Toronto Stock Exchange to allow junior companies to use analytical methods other than fire assay without forcing them to disclose, on an ongoing basis, results from checks using fire assay over the mineralized interval. Fire assay, apart from being a uniquely reliable extraction technique, uses small samples that readily reveal erratically distributed coarse gold.

Had this requirement been in place, Bre-X would have had to disclose results from both methods on duplicate samples. The erratic fire-assay results would have made it quite clear, early on in the game, that Bre-X’s sampling procedure warranted further investigation on the part of consultants, regulators and mining analysts.

— Next week: “Fire assaying doesn’t work on my samples.”