In the impression of many people, the process of mineral processing and drug addition relies on the experience of experienced masters. The teacher grabs a handful of potions, looks at the foam, and knows how much to add and how little to add. This method works in small mines, but it doesn't work in large mines.

Why? Because the scale of large mines is too large, relying solely on "hand grasping and eye watching" is no longer enough. Today we will talk about how the management of mineral processing agents in large mines has shifted from "relying on experience" to "relying on data".

Experience is unreliable

Experience is sometimes accurate, but sometimes it can deceive people.

For example, the teacher judges whether the dosage is appropriate from the color of the foam. However, the color of foam is affected by many factors - ore type, pulp concentration, inflation volume, lighting conditions, and even the weather of the day. The foam color of the same batch of medicine and the same kind of ore on cloudy and sunny days looks different. It is difficult to guarantee accuracy based on visual judgment.

For example, the master judge the purity of yellow medicine based on the hand feel. Take a squeeze and observe the speed of dispersal. But this texture is related to the moisture content, particle size, and environmental humidity of the yellow medicine. The same batch of medicine feels completely different in summer and winter.

Large mines consume a large amount of chemicals every day, and any errors based on experience will be magnified into tangible losses. So, large mines must abandon the thinking of "almost enough" and establish drug management on the basis of data.

Where does the data come from?

Building pharmaceutical management on a data foundation requires reliable data. Where do these data come from? From three aspects.

laboratory testing

The laboratory in the mine tests a large number of samples every day. Accurate data on the grade of raw ore, concentrate, and tailings must be obtained through laboratory testing. The concentration, purity, and impurity content of the drug should also be confirmed through instrument testing.

Without these data, engineers are like "blind men feeling the elephant" - they only know whether the indicators are good or bad, without knowing why they are good or why they are not.

online meter

Modern beneficiation plants are equipped with various online detection instruments. The pH value, oxidation-reduction potential, and drug concentration of the slurry can all be monitored in real-time. The operator can see the amount of each agent added and the trend of changes in various parameters of the slurry on the screen in the control room.

Although these instruments are expensive, large mines are willing to invest in them. Because they calculated that the medicine that the instrument could save them might be more expensive than the instrument itself.

Production Record

The operation records of each shift, every equipment adjustment, and every medication replacement must be truthfully recorded. These records are valuable materials for analyzing problems. The indicators suddenly deteriorated. By flipping through the records, you may be able to find the reason - it turned out that the previous shift had replaced a batch of new chemicals, and it turned out that the grinding fineness had changed.

Without records, there is no traceability. Without tracing, if there is a problem, we can only guess. In large mines, the cost of guessing is too high.

How to use data?

With data, one still needs to know how to use it. Data is not meant for storage, it is used for making decisions.

trend analysis

Data from a single day has limited significance. But by looking at data from one or two years together, patterns can be discovered. For example, the consumption of a certain medication will increase in a certain season every year. Analysis has found that the accelerated decomposition of chemicals is due to the increase in temperature. Then adjust the purchase plan in advance, strengthen inventory management, and eliminate the problem before it happens.

Association Analysis

The pH value of the slurry has changed, and the gold grade of the tailings has also changed accordingly. Is there any relationship between these two things? By retrieving historical data for correlation analysis, it may be found that the tailings gold grade is lowest when the pH value is within a certain range. Then make this range the control target and have the operators strictly follow it.

abnormal warning

When the online instrument detects that a parameter deviates from the normal range, the system will automatically alarm and remind the operator to handle it in a timely manner. Correct the problem before it expands. Avoid waiting until the indicators have deteriorated before making up for the situation.

The challenge of transitioning from experience to data

Transforming pharmaceutical management from relying on experience to relying on data is easier said than done.

Challenge 1: Trust Issue

Many experienced engineers have become accustomed to solving problems through experience after working for most of their lives. You make him believe in the instrument and the data, but he may not necessarily believe it. He would say, "There are times when the instrument is not accurate, but I can still see with my own eyes that it is reliable

To change this mindset, it takes time and patience. The best way is to let data prove itself - when data helps engineers solve a problem that was previously difficult to solve, they will gradually accept it.

Challenge 2: Equipment Reliability

Online instruments cannot be used simply by installing them. The slurry environment is harsh, and the instruments are prone to breakage, blockage, and drift. Poor maintenance and upkeep result in incorrect data. Wrong data is more terrifying than no data - it can lead people in the wrong direction.

Large mines require specialized instrument maintenance personnel to regularly calibrate, clean, and replace vulnerable parts. This job is tedious and unremarkable, but very important.

Challenge 3: Data silos

The laboratory data is in one system, the production data is in another system, and the inventory data is in a third system. Without connecting data, it is difficult to conduct correlation analysis.

To break through these data silos, system integration is required. This requires investment of funds, manpower, and coordination among various departments. Many large mines are doing this, but there is still a distance to go before it can be fully connected.

The benefits of data management

Is it worth investing so much effort in data management? The answer is affirmative.

Reduce medication consumption

By precise control of the medication, waste can be reduced. In the past, in order to prevent problems with the indicators, operators used to add more and more, and the dosage was often too high. With data support, we can find the "just right" usage and save the excess. Large scale mines require a large amount of chemicals, and although the savings are not significant, the absolute value saved is considerable.

Stable production indicators

Relying on experience to add medication results in significant fluctuations in indicators. By adding medication based on data, the indicators are more stable. Stability means stable recovery rate, stable concentrate grade, and smoother operation. The benefits brought by these 'stabilizations' are often greater than the savings of the medication itself.

Improve problem-solving efficiency

Previously, there were issues with the indicators, and it took a long time to investigate the cause - was it a change in the ore? Has the medication changed? Is there a problem with the device? With data support, many problems can be quickly located, and even online instruments can provide early warning. The processing time has been shortened, resulting in reduced losses.

Accumulate organizational knowledge

Remember the experience of the old master in your mind. When people leave, experience is taken away. The data is different, it is stored in the system. Even if there are personnel changes, historical data and analysis models are still available, and newcomers can continue to use and optimize them. This is the accumulation of organizational skills, not the accumulation of personal experience.

future direction

From experience to data, this path has been taken by large mines for a long time. But it's far from over. The future direction is deeper application of data.

Predictive control

The current control is more reactive - adjusting parameters when they deviate. In the future, we can achieve a "predictive" approach - based on the properties of the incoming ore, differences in batches of reagents, and changes in weather, we can predict in advance which direction the parameters will deviate and actively adjust the reagent plan. Eliminate the problem before it occurs.

auto-optimization

The current dosing system has a target value set by the operator, and the system automatically adjusts the valve opening to track this target value. In the future, the system can find the optimal target value on its own - continuously adjusting the target value based on real-time production data to make the indicators better and better. People only need to set the direction and boundaries of optimization, and the specific adjustments are completed by the system.

Knowledge precipitation

When data accumulates to a certain extent, useful models can be trained. What factors have the greatest impact on the recovery rate? When is the best time to add medication? What kind of chemical solution should be matched for ores in different mining areas? The answers to these questions can gradually be extracted from data and become knowledge assets of mines.

The transformation of large mines from relying on experience to relying on data is a quiet change. There is no grand scene, but every step is tangible. The core of mineral processing agent management is shifting from the "feel of an old master" to the "curve on the screen". This is not only a technological advancement, but also an upgrade of management philosophy.