Stage addition is common practice; it’s what being in control is all about. My personal preference is to add cyanide to the grinding mill, but this requires carbon in-columns to mitigate against spillage loss, generally 60-70% addition at the first point, and the remainder at no more than 2 top-up locations. But there is no point in multiple additions unless you are routinely monitoring cyanide and DO through-out the Leach circuit. In some plants what seems like a need for additional cyanide is actually a need for more air. You need to measure and understand the Leach kinetics.

Stage addition is a good way of controlling cyanide consumption, my experience on a CIP plant was that we got around 10% reductions in consumption rates. As provided you are already running a reasonable tail free cyanide concentration, you aren't too likely to see a recovery improvement, although this may depend on what cyanimides you have in the ore. Again, as said earlier, automatic control really makes a big difference here - the work done by guys feeding in to the ICMC shows that auto control can slash cyanide consumption, and therefore reduce the detox costs too.

Be careful, the saying “the more you add the more you use" has been proven by us and many more. A cyanide controller is a must for this line of thinking.
Residual cyanide at the end of the leach is a nice to have if the leach kinetics indicate no further dissolution and should be kept as low as possible for cost, capital and environmental reasons.

Get the pre conditioning Ph, DO levels and conditioning time right and downstream works a lot easier with high oxygen levels in the first three to four stages!

If you do not have a clear cyanide consumer in the slurry then use the ratio of CN-/O2 = 8. If not maintained one of the reactants will govern the dissolution reaction rate. Do a CN- and DO profiles of you circuit this will answer your questions of the leaching needs for cyanide and oxygen. Be aware that the indication of DO levels depends on the instrument type used.

Setting up a robust and effective automatic control is not as easy as it seems due to the need of reliable daily service routine on your analyzer (direct readout probes have mixed success), the response delays introduced by the size of the tanks you adding the cyanide to, relation between the cyanide added and pH etc. This is becoming important when you looking at a few % reduction in usage.

Adding cyanide to the grinding circuit speeds up the kinetics due to excellent mixing however the losses of CN to HCN can be fairly high and the danger of high HCN levels within the grinding circuit is significant therefore would not recommend it.

We were lucky in that our CIL feed was made up of a large number of sources thus resulting in a fairly average feed make-up (characteristics). The on-line analyzer was set at 5min / sample frequency and a very basic Ratio plus PPM feedback (PID) controller to add cyanide. Although a controlled / calculated split between the first and second addition point is ideal, we used a fixed manual split arrangement (2/3) on the dosing valves. Had results similar to (10%) reduction in consumption! As mentioned by others, no point in adding more than CN-/DO of 8, keeping an eye on final residue tenors and dosing splits of 2/3 between top and secondary stages.

Another good reason for staged addition of cyanide is that maintaining low concentrations of cyanide reduces dissolving copper minerals hence saving cyanide and yielding dore with less copper for less expensive refining.