1
24-053
Improved Drilling Controls Through Forecasting
Karl Ingmarsson
Boart Longyear, West Valley City UT
ABSTRACT
Larger mines spend millions of dollars annually on rock
drilling tools, but often purchase based on historic con-
sumption. The results vary from excess inventory to acute
tools shortages. These lead to emergency sourcing, some-
times mixing products that don’t perform optimally. Boart
Longyear analyzed the usage pattern of 96 thousand drill
bits at thirteen different underground mining contracts,
showing very high variability. The monthly standard devia-
tion on a product level was 42%, with the highest for
development and low volume bits. A new rule of thumb
is suggested to carry only one month of safety stock for
high volume bits, but increase to three months for low vol-
ume bits. This balance will reduce the risk of outages with
less inventory. The accuracy can be improved by consider-
ing the specific conditions at each mine and reclassifying
individual bits for a different safety stock level. Root cause
analysis and collaborative forecasting between mines and
suppliers can lead to further optimization of availability
and inventory levels.
PROBLEM STATEMENT
Measure-While-Drilling is used to predict downstream
impacts, but mine operators do not reverse the flow of geo-
mechanical knowledge and production plans to forecast
rock tools usage. Conventional wisdom has it that historic
consumption is a good enough predictor of future rock tools
needs. The average monthly consumption from the past is
used to forecast what is needed in the following months.
A common practice is to purchase one month’s expected
usage while having another two months’ worth of safety
stock. Theoretically, the inventory level will fluctuate for
between two and three months of usage. While this sounds
sufficient, we know empirically that mines often under
forecast. No mine is shut down because of a lack of rock
tools, but shortages impact both sourcing teams and opera-
tions. Emergency sourcing takes away from other priorities
and typically increase costs, while product substitutions or
mixing of brands can lead to poor drilling performance.
Boart Longyear decided to challenge the old rule of
thumb by analyzing reliable, actual consumption data from
mining contracts. To limit the scope, we focused on drill
bits since bit consumption has the highest degree of vari-
ability. The analysis is exclusively based on tophammer
drill bits used in underground metals mining. Leveraging
sound statistics, the objective was to identify better ways to
forecast, to manage inventory, and eventually drive longer,
more consistent bit life. Conceptually the logic is a proxy
for all rock tools and types of drilling applications.
BASIC METHODOLOGY
The starting point was to base the statistical analyses on as
much compatible bit data as possible.
• Thirteen underground mines in Peru, all metallic
and using tophammer drilling for both development
and production drilling.
• All sites under Boart Longyear tooling contracts.
• Mix of small to large operations, with tools con-
tracts ranging from less than US$250k to more than
US$2M per year.
• Identical reporting processes with monthly inven-
tory tracked by Boart Longyear.
• Planned constant production or amount of drill
meters.
The analysis followed the following steps:
1. Narrow the list to the regularly used bit designs.
This eliminated test, specialty, or rarely used bits.
It reduced the “high runners” to between two and
seven bit part numbers (SKUs) per mine.
24-053
Improved Drilling Controls Through Forecasting
Karl Ingmarsson
Boart Longyear, West Valley City UT
ABSTRACT
Larger mines spend millions of dollars annually on rock
drilling tools, but often purchase based on historic con-
sumption. The results vary from excess inventory to acute
tools shortages. These lead to emergency sourcing, some-
times mixing products that don’t perform optimally. Boart
Longyear analyzed the usage pattern of 96 thousand drill
bits at thirteen different underground mining contracts,
showing very high variability. The monthly standard devia-
tion on a product level was 42%, with the highest for
development and low volume bits. A new rule of thumb
is suggested to carry only one month of safety stock for
high volume bits, but increase to three months for low vol-
ume bits. This balance will reduce the risk of outages with
less inventory. The accuracy can be improved by consider-
ing the specific conditions at each mine and reclassifying
individual bits for a different safety stock level. Root cause
analysis and collaborative forecasting between mines and
suppliers can lead to further optimization of availability
and inventory levels.
PROBLEM STATEMENT
Measure-While-Drilling is used to predict downstream
impacts, but mine operators do not reverse the flow of geo-
mechanical knowledge and production plans to forecast
rock tools usage. Conventional wisdom has it that historic
consumption is a good enough predictor of future rock tools
needs. The average monthly consumption from the past is
used to forecast what is needed in the following months.
A common practice is to purchase one month’s expected
usage while having another two months’ worth of safety
stock. Theoretically, the inventory level will fluctuate for
between two and three months of usage. While this sounds
sufficient, we know empirically that mines often under
forecast. No mine is shut down because of a lack of rock
tools, but shortages impact both sourcing teams and opera-
tions. Emergency sourcing takes away from other priorities
and typically increase costs, while product substitutions or
mixing of brands can lead to poor drilling performance.
Boart Longyear decided to challenge the old rule of
thumb by analyzing reliable, actual consumption data from
mining contracts. To limit the scope, we focused on drill
bits since bit consumption has the highest degree of vari-
ability. The analysis is exclusively based on tophammer
drill bits used in underground metals mining. Leveraging
sound statistics, the objective was to identify better ways to
forecast, to manage inventory, and eventually drive longer,
more consistent bit life. Conceptually the logic is a proxy
for all rock tools and types of drilling applications.
BASIC METHODOLOGY
The starting point was to base the statistical analyses on as
much compatible bit data as possible.
• Thirteen underground mines in Peru, all metallic
and using tophammer drilling for both development
and production drilling.
• All sites under Boart Longyear tooling contracts.
• Mix of small to large operations, with tools con-
tracts ranging from less than US$250k to more than
US$2M per year.
• Identical reporting processes with monthly inven-
tory tracked by Boart Longyear.
• Planned constant production or amount of drill
meters.
The analysis followed the following steps:
1. Narrow the list to the regularly used bit designs.
This eliminated test, specialty, or rarely used bits.
It reduced the “high runners” to between two and
seven bit part numbers (SKUs) per mine.