DEAR ALL DURING A SUDDEN STOPPAGE OF CLOSED CKT BALL MILL I TOOK A ASAMPLE FROM I CHAMBER DISCHARGE (NEAR DIAPHRAGM) & I FOUND THAT IT IS PASSING 99% FROM 2MM SEIVE. WHAT DO U SUGGEST? GR. MEDIA TIPS WERE VISIBLE THROUGH OUT THE LENGTH OF CHAMBER. SHOULD I INCREASE SMALL MEDIA & REDUCE THE BIG MEDIA IN I CHAMBER TO INCREASE THE OUTPUT OR INCREASE THE MILL VENTILATION & PUT MORE FRESH FEED? IN II CHAMBER ALSO GR. MEDIA SURFACE WAS VISIBLE . RAJ
Michael Clark
It is quite common for the grinding media to be visible in the first chamber. Not so in the second chamber. The first step should be to increase the fresh feed rate.
BK
Dear sir,
Please check the following:
- Velocity across the chambers & it should not be more 1.5m/sec.
- Sound level across the first and 2nd chambers.
- Diaphragm condition i.e,. blockages
- Percentage of grinding media in first and second chambers.
- fresh feed size etc,...
by Bandari.K
STEAG Powitec - Albrecht Keiser
Use a closed loop ball mill optimiser, predicting the blaine and controlling it to the optimum.
Concept:
Integration of additional sensors
(microphones at the mill and vibration sensors at pipes and separator)?
Automatic feature selection and extraction (significance ranking)? of existing process data
Automatic model generation (regression, neuronal networks, probabilistic nets, Gray-Box-Models)?
Set point integration into the DCS/PCS
Technical Advantages:
Online monitoring of separator(s) with vibration sensors to allow for Blaine prediction (Soft Sensor) by adaptive and non-linear Neural Networks and quality based throughput maximisation
Recipe independent closed-loop control by using sound and vibration features, Blaine prediction and Advanced Process Control (Bayes controller)
2 microphones enable proportional measurement and allow for advanced signal processing; wired connection delivers high signal density
Rugged construction: Metal housing and metal membrane. No field computer
Data pre-processing of all relevant
process data by analysis of correlations
Advanced signal processing on acoustic measurements
Visualisation highly flexible (integration of PCS data);
Graphical User Interface with trending and alarming
Modular system: Upgradeable to Kiln-Optimiser,
easy and inexpensive to scale to multiple mills
Results:
Reduced quality deviation (Blaine) by up to 30%
Between 2% and 6% less specific energy consumption
Up to 10% more throughput
Full automatic closed-loop control
Recipe independent throughput maximisation
Self learning of controller parameters for different recipes
No re-parameterisation at new recipes
Elimination of lab delay
Fast adoption to change in clinker types
Contact: a.hanf@powitec.de