Custom Hardrock Milling and Concentrating

Welcome To The Canadian Gold Prospecting Forum. Please consider registering to gain full access! - Its free and only takes a few seconds to complete - Click Here.

Gold Prospecting Forums, Gold Prospecting Forum

[Guest]

A A A A

[mbmmllc]

Hello all-

   I wanted to post to let you know about my new ideas and equipment I have put together for custom milling and concentrating hardrock ores.  Some of you may recognize “popandsonminers”: I’m the “son”.  I am certainly interested in getting exposure and new customers and some of you may be interested in my services.  However, I would also like to get feedback from the experienced miners out there who can give me some insight into things I should try and things to avoid.

   I have been working for the past four years in a small underground gold mine.  We produce around 500 tons of ~½ oz/ton ore each year which gets shipped ~250 miles away to the refiner.  The large expense of shipping has been eating up most of our profits.  To solve the problem, I put together a small sampling mill to determine if it was feasible to mill on site and recover a high percentage of gold.

   I get 90% gold recovery using only gravity by crushing the ore to 90% minus 80 mesh.  These recoveries are only possible by capturing a large percentage of the sulfides in the ore (which carry between 10% - 50% of the values).  This summer the mine owners are going to do some bulk sampling (10-20 tons) on-site to see if these high recoveries can still be achieved on a “production scale”.  If the bulk sampling can accomplish the same high recoveries the mine owners are going to scale up to a mill that can process 1-3 tons/hr.
   
You can view my videos of our underground operation and crushing/testing equipment by searching for “mbmmllc” on YouTube, and you can visit my website at Mt. Baker Mining and Metals and read more about what we do.  

Thanks,

Jason Gaber
Mt. Baker Mining and Metals
mbmmllc@gmail.com

[aumbre]

I’m excited and encouraged by your progress and results so far. You are doing exactly what I would like to do on my property in Oregon. Bulk testing such as you are doing is probably the only way to evaluate the potential for small operations, especially when considering ore from “spotty” veins and different hand sorted classes from old dump piles or tailings.

[mbmmllc]

aumbre-

What type of testing have you done on your vein system?  What I have found with my testing is a chip assay across the vein is fairly consistant (2 taken from the same spot on the vein) if done right (there are many posts and articles on how to take a correct assay), but can be vary tremendously 5'-10' away. 

For example in one of our stopes we took an assay every 5' for 30' and our assays are as follows (all in opt) .317, .753, .094, 1.45, .550, .224.  These assay were taken by an experienced geologist (30 years experience).  We have taken hundreds of assays both of our vein system and our muck samples and we average between .5-.75 opt.  I have come to accept that many many assays are needed before a true "average" can be determined, but for a small operation most of the time the $'s aren't there for hundred+ assays if an assay costs between $20-$50.  This just goes to show you the many "hidden" costs of mining or how expensive it can really be, because most people don't consider assaying to be a all that expensive, however our operation spends 1000's of dollars a season on assays.

Another topic that I will bring up here is the variability in the vein width.  In our vein system the values are mostly concentrated along the footwall.  Most of the time free gold can be found in the face up to 3" off the footwall after washing down a fresh blast.  We have taken samples and found that up to three inches off the footwall the grade runs between 5-15+ opt, while the rest of the vein is mostly less than .25 opt.  So in reference to your previous post, I absolutely agree with you: veins and most deposits are spotty/variable and thorough testing and sampling should be done before any large investment is made in a property.

TEST, TEST, TEST... it may seem like a pain and a waste of money, but if it is not done, or not done CORRECTLY then the cost can be severe and may cause a good property/project to fail.  For example the Alaska Gold Rush guys on TV spent how much money and time for how many ounces?

Here is something else I will mention since I have the time.

When I first started conducting the testing of my mill my sample plan was as follows:

Head ore sample taken after the roller mill (-1/8" material) by holding a container under the stream for 1 second every 20 seconds.

Concentrate sample was whatever ended up in the concentrate bucket (for a 50 lb of sample)

Tailings sample was taken at the discharge of the hydrocyclone (-80 mesh) by holding a container under the stream for 1 second every 20 seconds.

I don't think I had one sample out of 10 (total of three assays: heads, cons, tails per sample) that would add up.

An example of one of my samples is below:

Head Sample: .667 opt

Cons Sample: .411 opt (after adjusting for weight of con and head weight, essentially how much of the values were recovered)

Tailings Sample: .091 opt

The tailings and cons sample should add up to the head sample right?  But they don't, not even close in percentage.  I was starting to get very frustrated and confused.  Was I not sampling correctly or were assays just a bunch of voodoo?  I finally called the mill manager at Kinross and talked with him about my process and he the correct way to conduct sampling of this sort is to assay the cons and assay the tailings and to back calculate the head sample.  So for the example above the calculated head sample is

.441+.091 = .532 opt

I have read a few different hardrock mill testing papers and they come to the conclusion the head sample must be back calculated because the head ore is too variable to get a representative head sample.  Does anyone have any additional thoughts on this?  Mr. Science, what kind of mill testing regime do you use or have heard used?

Before I go I would like to thank everyone for their support and encouragement about my website and youtube videos.  I hope I have given some of you ideas about how to proceed with your projects.

Thanks again

Jason Gaber
Mt. Baker Mining and Metals
mbmmllc@gmail.com

[aumbre]

Jason, in answer to you question about sampling, I’m still in the investigative and learning mode and I take three types of samples depending on the situation. I guess that I’ll explain a little about our property and what I’m trying to do. The mine on our property was started in the late 1800’s ,has about a mile of drift on the vein on four levels, and was heavily stoped above three levels. The vein width was usually about three or four feet wide and average grade mined was about 1/3 ounce per ton gold and a little silver. There is still some ore left but I’m not expecting a Bonanza. I’ve only taken one “grab sample” from the dump pile- 0.07 Au and 0.54 Ag oz. per ton. My partners took some samples in the early ‘90s selected for visible sulfides that carried about 0.3 Au and 2.0 Ag ounces per ton with over 1% each of Cu, Pb, and Zn, and about 10% Fe.
An example of the sulfide ore.


Some partially oxidized ore.

There are scores of other ledges or veins on our property that I have been exploring, most are surface exposures but some have had considerable underground exploration. On these exposures I’m sampling and mapping, looking for pockets and/or indications of possible oreshoots. It appears that most of the cuts were put in on vein intersections and with all the different types of mineralization brecciazation, discontinuities, etc it is sometimes difficult to tell what I’m really looking at. If there appears to be a well defined vein I’m taking “chip samples” across a measured width. Most of the sampling that I’ve done over the last two years has been what I call “character samples”. I guess these are really “grab samples”, I’m picking out little seams of mineralized rock or picking up the most interesting pieces off the little dumps for analysis, realizing that these are nowhere near a representative sample, but if the assay is encouraging it indicates where more work is warranted and the types of mineralization to be on the lookout for.  

[aumbre]

I was pushed away from the computer before I was quite finished last night. I wanted to show one type of cut that would be a candidate for some bulk sampling. There is a somewhat complex fracture system(s) going on here and the mineral values seem to be pretty erratic.
In 2009 I picked out a small amount of vein material for assay and was surprised when the results came back with 3.25 Au and 20.65 Ag. (All result in ounces per ton.) I assumed the values were in the reddish tinged galena material associated with vuggy (amethyst) quartz.


I returned in 2010 to clean out the old cut and take more samples.A picture from a different angle showing the floor of the cut and preparing to take a chip sample.

A 32 inch chip sample being taken parallel to the ribbon. ) 0.179 Au and 2.04 Ag.

Grab sample from material to the right of the chip sample. 0.088 Au and 3.56 Ag.

A sample of what I thought would be “highgrade” was hand picked from existing broken rock found in the cut assayed 1.85 Au and about 2.0 Ag.
So far I’m out about $200 for assays and have no gold in my poke. Would this be a good location to “bulk sample”?

[mbmmllc]

Aumbre

First.  Great pictures, it is much easier to determine what a person is talking about with a good picture.

Second.  I think you are doing exactly what you should be doing in your situation (in my opinion).  You are doing your sampling and mapping, becoming familiar with your workings/ore type/grade etc. without breaking the bank. 

It sounds to me like your strategy is as follows (just to clarify and reiterate):  Sample and map to determine average grade and ore character.  Determine what “high grade” looks like and identify places in your workings that may contain high grade (i.e. brecciated portions of the vein, vein intersections, Jogs or turns in the vein, etc.).  Again this is exactly what I would be doing in your situation. 

My next question for you is:  What are your goals for the property and what do you think your next step may be?

I am going to make an assumption here and assume you plan to open up some of the old workings to access the vein and begin small scale mining again with a small crew of your own or hire someone to work for you.  The other option you may be thinking about is proving up the property with samples and maps and then trying to promote/sell/lease the property to another company.

Lets say you are going to open up the old workings and begin mining and milling on site on a scale of 1-10 tons/day.  If you have found a higher grade portion of the vein which has some potential to produce tons this is where I would start.

I am going to take a little side trip here and talk about mining costs.  I work for a small mining company with 3 employees and I will try to summarize our mining costs to get you in the ball park to determine if your ore is of sufficient grade to warrant mining.

We have a 3 man crew working 12 hour days 6 days a week.  We average somewhere between 60-90 days/year due to snow pack late in the summer and snow early in the fall.  Our mining costs are approximately $120,000/season or around $2,000/day.  This is just our mining costs this does not include trucking, crushing, milling etc.  We produce around 500 tons/year, which brings our mining costs to $240/ton.  I think an average milling fee for on site milling would be around $150/ton.  So to review lets say it costs $250/ton to mine $150/ton to mill and make a con, and another $100 per ton (of bulk ore produced not $100/ton to refine the cons) for final refining to produce a saleable product.  Our total mining costs are 250+150+100=$500 per ton plus or minus. 

Also keep in the back of your mind the milling and concentrating process will only recover, let’s say, 90%.

This means we need at least $550/ton of values to break even (considering 90% recovery).  So during your mapping and sampling process keep these numbers in mind.

Now I don’t want to discourage you, I just want to make sure we all (not just aumbre, but everyone reading this) keep in mind the costs involved in small scale mining.  These prices may vary depending on the site, width of vein, mining method etc, but this is a good starting point for getting something down on paper.  (There is another good post here called “hard rock production costs” which has loads of great info about mining costs)

I am getting way off track here and will come back to your original question: “Would this be a good location to bulk sample”.  I would say that if you think this is an area with potential for ore development, the ore grade is promising (i.e. $550+/ton), and there is enough reserves to warrant starting a mining operation then, yes this would be an area to bulk sample. 

Even if this area doesn’t meet one of the above criteria I would still say consider bulk sampling.  Then you will know if the ore is amenable to gravity concentration or does it need to be shipped off to a mill hundreds of miles away (huge costs).  Most big mills with need a minimum amount of ore before they will consider processing (1000 tons is a number I have heard for Kinross’s Kettle River mill in Republic, WA).  So bulk sampling will answer some key questions for you in how you proceed with your project, even if you are still on the fence about other questions.

A few questions for you:

Have you tried hand crushing and panning to determine free gold?

I have used to following tests for free gold in the past on ore samples:  If you can determine what high grade ore looks like, take a sample of say 5-10 lbs of high grade ore.  Weigh the sample accurately (to a 1/10th of a lb.) and hand crush the rock to a powder (it takes a while, but it’s worth it).  Weigh the gold that’s recovered and using the weight of the gold and the weight of the original sample you should be able to determine (very roughly mind you) your oz/ton in that sample.  Now that is a long explanation to make this point.  If your high grade samples assay 1-5 oz/ton and you can only pan .1-.2 oz/ton from the sample then it doesn’t look good for gravity concentration of free gold.  However, this doesn’t mean you couldn’t make a gravity concentrate of the sulfides and refine them to extract the gold, it just means your ore does not carry a high percentage of free gold.  You could even expand the experiment to assay and screen the panning tailings (crushed quartz) and assay the panning cons (sulfides and gold) and this will tell you something about liberation and percent recovery if the sulfides are removed from the quartz.  These are all very rough calculations and should be used again to get a ball park figure.

What does the old literature say about your mines history? What grade did they average?  What vein thickness?  Where there ore shoots and barren areas or was the material fairly consistent in grade?

The old mining records/literature can be invaluable information when trying to open up an old mine.

Do you have an estimate on the potential reserves left in the current workings?

I think I have rambled on plenty for one post and hopefully have given you something to stew on for a while.  I think you project definitely has potential.  If the old timers worked it to the extent you say, there must have been something worth getting.  Keep us posted and I would very much enjoy to hear about anyone else’s projects they have going or old workings they own.

Jason Gaber
Mt. Baker Mining and Metals
mbmmllc@gmail.com

[aumbre]

Jason, thanks for the great reply with lots of useful information. You pretty much hit the nail on the head concerning my sampling/ mapping plans on the undeveloped veins. The old mine has some potential but as I said I am not expecting the Big Bonanza. Until the 1940s the mine consisted of three levels about 120 feet apart with the longest, lowest (haulage) level going back about 1800 feet. The ore occurred in shoots mostly about 100-300 feet long and 50 to 250 feet vertically. It is estimated that about 30% of the vein material above the haulage level was removed, mostly by overhand stopes with stull timbering. In the early days it was a profitable mine, with some portions assaying 1 to 2 ounces gold per ton, but over the life of the mine (about 100,000 tons?) the average recovery was probably about 1/3 ounce per ton. In the upper levels there was some stoping on parallel veins that are separated by 3 to 10 feet of altered rock. Numerous crosscuts somewhat explore the parallel veins which kind of peter out or lose their identity on the lower level.
I’m not planning on opening the mine anytime soon, although it wouldn’t be too hard (surface slides across the entrance) and most of the haulage level is in pretty good shape. I was back to the end in 1981.
The $500.00 per ton figure is very useful as a general guideline for smaller hardrock operations. Unfortunately much of our ore is not exactly free milling and it doesn’t appear that there is a ready market for sulfide concentrates that will pay for all the precious and base metals contained.
Well that’s enough for now, I’ll get back soon with some more info and perhaps my plan on milling sub economic waste rock ( dump) piles.

[aumbre]

Jason, I reviewed the chapter on ore sampling in “Peele’s Mining Engineer’s Handbook” on ore sampling and could find no mention of the problem you encountered with sampling the head ore. Up until about the 1940s there were many custom mills, smelters, ore buyers, etc that would buy smaller lots based on head samples and though the buyer’s and seller’s assays didn’t always agree they were close enough for transactions to be completed. The book illustrates industry approved sampling methods, general considerations, and gives examples of several ways that errors could be introduced into the process. Some of the potentials for error include inaccurate weighing (including moisture), insufficient crushing and mixing, and methods that are not taking representative samples of the various size fractions in a segregated ore.

[mbmmllc]

Aumbre and others

We may be trying to compare apples and oranges and I think the following example may help, here it goes.

Our mine has a contract with Kinross’s Kettle River Mill in Republic WA.  We send our raw ore to Kinross where they use flotation and cyanide to extract the gold values.  Over the last three shipments they have managed to extract between 96%-98% of our values (according to lab tests).  The mill has a capacity of around 1500 tons per day.  They run their ore through the mill continuously and have a large fine ores bin which has a two day holding capacity (~3000 tons) which feeds their ball mills.  When our 500-1000 tons of ore has arrived at their site they shut down their coarse grinding circuit which is composed of a jaw and cone crusher and crush all of our ore to minus 1” (it takes about 4 hours to crush our whole seasons work!).  They take a sample off the end of the conveyor every 5 minutes using a shovel to catch the falling ore.  At the end of the test between 200-300 lbs of ore has been pulled out of the stream which is what we use to determine head grade.  They use this method so they can keep running their ore from their fine ore bin through the mill while they sample our ore.

I wrote all the above to explain the following:  Kinross does not shut down their entire mill to run our 500 tons and get our gold separately.  They mix our ore in with their own ore so the milling process does not determine how much gold comes out of our ore, this must be done using laboratory tests.

To determine percent recovery Kinross splits the head sample mentioned above into 3 subsamples.  Kinross gets one, we get one, and the third is kept our for an umpire assay.  Part of Kinross’s sample is sent to their lab where the following happens. 

They crush the sample to 80% minus 280 mesh and bottle roll the sample in a cyanide solution for 5 days (the retention time in their cyanide tanks in their milling operation).  After 5 days they determine how much gold was dissolved in the cyanide and assay the bottle roll tailings.  By taking the two numbers (amount of gold dissolved, and amount of gold in tailings) they calculate the percent recovery.  They do not use the head ore sample to calculate the percent recovery. 

So for this example lets say Kinross determines through their lab tests they can recover 96% of the gold.  Kinross sends head ore samples off to an independent assay lab, and we do the same.  If the assays are within 10% of each other the samples are averaged together and that is the amount of gold we get paid for (taking into account total dry weight of the entire pile).  If there is more than a 10% difference then the third sample (umpire sample) is used and which ever number the umpire assay is closer to those are the two numbers used to determine head grade.

So for example if we sent them 600 dry tons of ore and the head ore samples came to .410 opt (Kinross) and .396 opt (our sample) the average of the two samples is .403 opt.  .403 opt*600 tons=241.8 ounces.  If Kinross determines they can recover 96% then we get paid for: 241.8 ounces *.96= 232.128 ounces.  That is how our contract works with Kinross.

As a side note Kinross’s fees are as follows:

$65/ton milling fee

5% head assay value

Consumable fee (cyanide, lime etc.)  approx $10/ton

So Aumbre you are absolutely right about how the head grade is determined for a custom milling process and it is still more or less the same today as it was in the 30’s and 40’s.

Now to talk about the method used for determining percent recovery in gravity testing.  I started a few years ago trying to determine the feasibility of gravity recovery on our ore and my boss wanted to know what percent recovery I could get.  I tried a few things and my numbers would never add up.  I emailed Barney Darnton (the mill manager at Kinross’s mill, since retired) and explained my problem to him and asked him what to do about it.  His response:

“At a research facility… they would use the con and tail weights and respective assays to calculate the head and the recovery.  For a gravity test it is universally acknowledged and accepted that the feed grade is the sum of the products and the recovery is calculated as opposed to measured.” 

Now to this day I can’t explain why this is and why I can’t get the con and tail assay to add up to a head assay.  It may be the samples have too much free gold in them and the head or con assay gets thrown off by the “nugget effect”.  I can tell you I have run approximately 20 tests assaying the heads, cons, and tails, and only on rare occasion add up, most are wildly off.  I have been using Barneys method ever since and to be honest with all of you I don’t know how reliable or accurate it is.

Here are a few links to papers discussing gravity recovery methods.

http://www.knelsongravitysolutions.com/sites/knelsongravity/files/reports/report21s.pdf

note on page #3 of this paper under “BASIC RESPONSE” they state “overall recovery is based on the assays of the three concentrate samples and the tails…”

http://www.knelson.ru/sites/knelsongravity/files/reports/report20s.pdf

http://www.sid.ir/en/VEWSSID/J_pdf/8542008B102.pdf

There are lots more papers out there too.  Most seem to be using centrifuges units for their concentration. 

I hope this helps, I have tried to learn as much as I can from the experts and veterans in the industry and sometimes I don’t completely understand the reasoning behind it.  Only if I had an unlimited budget and a fully equipped lab to really flush out this whole problem… maybe some day.

Thanks everyone for reading

Jason Gaber
Mt. Baker Mining and Metals
mbmmllc@gmail.com

[aumbre]

So, Kinross pays on assay results from your head ore- not by “back calculating heads and tails”. The sampling method sounds like it could easily be inaccurate when collecting the first split and that inaccuracy would follow down to all the remaining splits. It’s their mill and I guess that they can sample anyway they please, but I would be wondering what the assay might be if the sample was collected by a standard recognized method.