01/12/2015 Energy efficiency improvement in a hot strip mill by using variable speed drives EE Publishers
EE Publishers (http://www.ee.co.za)
Energy efficiency improvement in a hot strip mill by
using variable speed drives
A hot strip mill (HSM) contributes a significant component to the overall electrical load of a modern steel
plant
Primary processes in a HSM that are directly accoutable to the production of the end product include the large
electrical motors that drive the mills.
Secondary processes that are not directly accountable to the production of coils include large descaling water
pumps and interconnected water systems that are involved in the treatment of the process cooling water and
descaling water. The majority of these secondary processes have large water pumps with bypass facilities, resulting
in energy wastage during water bypass periods. The current inefficient way of operating these secondary processes
initiated an investigation into means to improve their energy efficiency.
In a modern steel plant, production, from iron ore to hot rolled coil, is achieved in as little as 16 hours. The typical
steel plant consists primarily of three business units: Iron making, steel making, and mills/rolling, as is
schematically shown in Fig.1. These business units contain energyintensive processes, and consume large
amounts of electrical energy and fuel gas per annum.
In the iron making business unit, the air separation unit and iron making processes are the main energy consumers.
Fig. 1: Schematic layout of a modern steel plant.
In steel making, the electrical arc furnace is the largest single consumer of electricity, typically contributing 30% of
the plant’s total electrical energy consumption. In the mills/rolling business unit, the largest consumption of
electricity is in the hot strip mill (HSM), typically contributing 10% of the plant’s total electrical energy consumption.
Modern legislative requirements dictate that plants must endeavour to reuse all water to have a minimal impact on
the natural water sources. The water treatment plant (WTP) combines various treatment and recycling processes to
minimise water purchases. The WTP typically contributes 8% of the plant’s total electrical energy consumption.
This paper will introduce energy efficiency improvement opportunities at the secondary processes, focussing on the
water systems that service the HSM. Varying monthly production figures at the HSM will have an impact on the
monthly energy savings targets to be achieved under the DSM contract. The monthly DSM performance targets are
therefore linked to plantspecific variables. The mechanism of an energy savings performance curve of which the
boundaries are determined by these plantspecific variables will be discussed briefly.
Hot strip mill – an overview
The roughing mill (RM) and finishing mill (FM) are primarily responsible for production in the HSM.
Fig. 2: Schematic layout of hot strip mill.
Slabs of various widths from the Continuous Caster are milled in the roughing mill (RM) and then in the finishing mill
(FM) at the HSM. Slab widths typically vary from below 1,000mm to about 1,600mm, with thickness between 75
http://www.ee.co.za/article/ist15406energyefficiencyimprovementinahotstripmillbyusingvariablespeeddrives.html 1/6
01/12/2015 Energy efficiency improvement in a hot strip mill by using variable speed drives EE Publishers
and 90mm. Fig. 2 shows a schematic layout of the HSM.
In the roughing mill the “slab” becomes a “transfer bar”. A transfer bar is coiled in the coil box between the RM and
FM, and uncoiled before it enters the FM where it is rolled to its final thickness. There are descalers at the entrance
of both the RM and FM for quality control purposes. The descalers ensure that scale that forms on the surface of the
material during milling is not milled into the material. Descaling is done by forcing water at a high pressure (typically
200 – 230 Bar) onto the material, forcing the scale away from the steel.
The RM and FM rolls are cooled with process water during the milling process. Water is supplied at typically 10 Bar
to the mill cooling headers.
At the exit of the FM, a runout table (ROT) with laminar cooling reduces the temperature of the strip before coiling in
the downcoiler.
Fig. 3: Typical timing diagrams of the mill process.
The utilization of these secondary water processes is linked to the milling process at the HSM. No descaling takes
place at the RM or FM descaler station if no milling takes place at the particular mill at that time. Similarly no roll
cooling at the RM or FM takes place when that particular mill is not milling. And when there is no strip on the ROT,
no Laminar cooling takes place either.
Fig. 3 shows a timing diagram of the milling process in the HSM. At low production rates – fig. 3(a) – there is a
relatively large gap between RM and FM milling (“Gap1”), and also between FM and RM milling (“Gap2”). The total
cycle is relatively long. The duration of milling in the RM and FM can be described by normal distributions. The
duration of the gaps between them will be determined by the production rate of the plant. Figure 3(b) shows a timing
diagram when the plant has a faster production rate than in 3(a).
The production rate of the plant will determine if simultaneous milling will occur in the RM and FM. If the capacity of
the upstream and downstream processes is adequate, it might happen that a slab is being milled in the RM at the
same time as a transfer bar is milled in the FM, as is shown in Fig. 3(c). In this illustration the “Gap1” duration is
zero and “Gap2” duration is shorter than in Figure 3(b). The cycle time is relatively short. In this case the secondary
water processes will be utilised much more than for the case of a low production rate as is shown in Fig. 3(a).
For a specific steel plant, the maximum hourly production rate will be determined by the peak production capacity of
the bottleneck process in its production line.
Other factors that will influence the monthly production rate at a HSM include the grade/quality of the material that
has to be produced (i.e. order dependent), the number and duration of operational delays (i.e. operations dependent),
and the number and duration of unplanned stoppages.
The average hourly production rate will affect the energy savings potential on the secondary processes of the HSM,
as will be shown in a later section.
Energy saving opportunties
Descaler pumps – case study: 5 highpressure pumps (driven by 11 kV motors rated at 2,800 kW each) are
installed, of which 2 run continuously to maintain 200 – 230 Bar pressure for use at the roughing mill and finishing
mill descalers. Three are on standby. Fig. 4 shows a schematic layout of this water system.
The water demand at the RM descaler is larger than the water demand at the FM descaler due to the number of
nozzles installed at the RM. Two descaler pumps need to operate simultaneously to deliver water to the RM
descaler, while only one pump is required to deliver water when the FM decaler is in operation. However, currently
http://www.ee.co.za/article/ist15406energyefficiencyimprovementinahotstripmillbyusingvariablespeeddrives.html 2/6
01/12/2015 Energy efficiency improvement in a hot strip mill by using variable speed drives EE Publishers
two pumps run continuously. Operational delays occur (e.g. changes of rolls on the mills) during the milling process.
During delays the discharge valves are closed. However, the descaling pumps continue to operate and water is
bypassed back to the scale pit, and energy is wasted.
For bypass purposes each pump is equipped with a bypass line with an orifice to ensure the minimum required
bypass flow. All bypass lines are currently equipped with a controllable valve that is closed automatically when the
pump delivers water to the descaler headers.
The power consumed by the descaler pumps can be decreased by the use of variable speed drives (VSDs) to stop
and start pumps during operational delays (currently they are not stopped due to damage caused to the motors by
frequent starting. Damage occurs due to the high starting currents which are eliminated when using variable speed
drives); the use of VSDs to reduce the speed of pumps during the gap times between RM and FM descaling periods
(currently the pumps run at fixed speed and water is bypassed at the pumps);
Only when RM descaling is on are two pumps needed to operate to deliver the high flow at ~230 Bar pressure. Use
of a VSD allows one pump to be stopped when the RM descaling period is finished.
Fig. 4: Water system of descaler pumps at HSM.
Fig. 5 shows typical “before” and “after’ profiles of water flow [in m3/h] and total electrical power [in kW] of the
descaler pumps. In terms of water flow, the same amount of water as before will flow during RM and FM descaling
periods when VSDs are used to control the speed of the pumps. Outside the descaling periods (i.e. during the gaps
and delays) less water will be bypassed back to the Scale Pit and energy can be saved.
Fig. 5: Descaler pumps flow and power profiles.
In terms of power consumption, during the RM descaling period the same power as before will be used with the
VSDs since 2 pumps at the same speed as before will still operate to provide the required high flow at high
pressure. During the FM descaling period only one pump has to operate, therefore a reduction in power can be
expected. During gap times and delays only one pump at lower speed has to operate, therefore a reduction in power
can be expected.
Mill cooling pumps – case study: Warm process water returns from the descalers to the scale pit. Water is pumped
from the scale pit through a filter system to a cooling tower, where the cooled, clean water is stored in the cooled
water basin, as is shown in the schematic diagram in Fig. 6. Clean water is supplied from the water treatment plant
(WTP) to the descaler pumps at low pressure by five 3Bar pumps.
From the same cooled water basin clean water is supplied by five 10 Bar pumps to the mill rolls for cooling
purposes. When no cooling is required, the control valves at the mill stands close and water is bypassed at the
WTP back to the scale pit, and energy is wasted. When VSDs are used on the descaler pumps, less water will be
bypassed at the descaler pumps during the gaps and delays (see Fig. 5). It will directly influence the amount of
water to be supplied by the 3Bar pumps. The power consumed by the 3Bar pumps can be decreased by using
variable speed drives to control the speed of the pumps by taking the pumps’ discharge pressure as a controllable
variable. If demand for water drops, discharge pressure will rise above the setpoint of 3 bar, and pump speed can be
reduced to meet the lower demand at the set pressure of 3 bar.
The power consumed by the 10Bar pumps can be decreased by using variable speed drives to control the speed of
the pumps in order to reduce/eliminate the bypass of water at the WTP during delays or when the demand for
cooling water drops at the mill. The speed of the pumps can be controlled by taking the pumps’ discharge pressure
as a controllable variable.
http://www.ee.co.za/article/ist15406energyefficiencyimprovementinahotstripmillbyusingvariablespeeddrives.html 3/6
01/12/2015 Energy efficiency improvement in a hot strip mill by using variable speed drives EE Publishers
Clearly by reducing the amount of water being bypassed back to the scale pit, the amount of water that has to be
moved by the ten scale pit pumps through the filter to the cooling tower will be reduced accordingly. This will reduce
the average number of scale pit pumps that has to be in operation over time.
The cooled water basin has a relatively large storage capacity that will smooth out the dynamic nature of the water
flow in the system. This will reduce the frequency at which the scale pit pumps will have to switch on and off.
Therefore the use of soft starters on the scale pit pumps for energy saving purposes will be a feasible solution, and
remove the need to use more expensive VSDs on the scale pit pumps to achieve the potential savings.
Fig. 6: Mills cooling water and decaling system.
Laminar cooling pumps – case study: A total of five pumps are installed to feed water from the cooled water
basin to the overhead storage vessel at the HSM laminar cooling process. Four pumps are normally in operation,
and one on standby. These pumps run continuously, resulting in overflow of the storage vessel back to this water
system’s scale pit. Fig.7 gives a schematic layout of this water system that serves the HSM.
Headers with spraying nozzles on the runout table after the finishing mill provide for cooling of the strip before
coiling. The valves to these headers open when the strip passes over the runout table, and closes when no product
is present on the runout table. The water to these headers is gravityfed from the overhead storage vessel. When
the valves close, the four feed pumps continue to pump and water simply overflows in the vessel back to the scale
pit, and energy is wasted. A total of five scale pit pumps are installed to pump water from the scale pit via filters to
the cooling tower. Four pumps are normally in operation, and one on standby. The motors run fixed speed at all
times. During shut downs the motors are stopped. The level of the cooled water basin determines the number of
scale pit pumps that is in operation at a time.
Fig. 7: Laminar cooling water system.
The power consumed by the laminar cooling feed pumps can be decreased by using variable speed drives to control
the speed of the pumps in order to control the water flow to the overhead vessel and reduce/eliminate the overflow
of water at the vessel for periods when demand for laminar cooling water drops at the mill. The speed of the pumps
can be controlled by using the overhead vessel’s level as a controllable variable.
The power consumed by the Scale Pit pumps can be decreased by reducing the amount of water being bypassed
back to the scale pit, the amount of water that has to be moved by the scale pit pumps through the filter to the
cooling tower will be reduced accordingly. This will reduce the average number of scale pit pumps that has to be in
operation.
Energy saving analysis
The previous section illustrated by means of actual case studies that energy saving opportunities are present at
secondary water processes of a Hot Strip Mill. The energy savings that can be realised from all these cases that
were discussed is directly a result of the method of operation of the primary process (milling/rolling).
It was explained by means of Fig.3 that the “onoff” batchlike nature of the milling operation will influence the
utilization of utility water for descaling and cooling purposes at the mill. It is this batchlike nature of the mills that
presents an opportunity to stop or slow down water pumps instead of bypassing water back to the sumps.
To determine the economic feasibility of a project of this nature to retrofit motors with VSDs or soft starters and
integrate the control of these systems into the plant's existing control system, the potential energy savings and
project costs have to be quantified. These figures will then be used to determine the economic feasibility of the
http://www.ee.co.za/article/ist15406energyefficiencyimprovementinahotstripmillbyusingvariablespeeddrives.html 4/6
01/12/2015 Energy efficiency improvement in a hot strip mill by using variable speed drives EE Publishers
project.
The important plant variables that have to be processed and analysed at a hot strip mill in order to quantify the
potential energy savings in the secondary water systems, include:
The operational schedule. From historical data a distribution of mill duration for the various mills must be
determined, from where a 12month average can be determined. Also the 12month average gap times between mill
operations should be determined.
The operational delays. From historical data a distribution of delays must be determined, together with a 12month
average.
The production rate. From historical data adistribution of the monthly and daily production rate must be determined,
together with a 12month average.
The range (mix) of products. If a range of products is normally produced (e.g. coils of various width and thickness)
that may influence the distribution of the variables listed above, then the sensitivities have to be analysed to
determine the influence of each of these product types on the above variables. Using 12month averages on these
variables should filter out any monthtomonth variations in the mix of products to be produced.
Planned shut downs. The duration of planned shut downs per annum must be determined and subtracted from
calendar time to give the time available for production. During shut downs no energy savings can be achieved on the
secondary water systems because the pumps are already stopped during these periods.
Water flow and pressure. The current water flow and pressure per system must be determined. Historical data from
the plant’s SCADA should be available to create trends that can be analysed. If not, then measurements must be
taken during a measurement audit.
Motor electrical power. The current electrical load profiles per motor under investigation should be obtained from
historical data or measured during a measurement audit. Alternatively, when water flow and pressure profiles are
available from e.g. a SCADA system, the motor power can be calculated by using a mathematical model of the
motor/pump combination, using pump curves and rated motor efficiencies.
A comprehensive energy savings model is then used to calculate the potential energy savings for a given set of
parameters. The parameters will include the various average time durations of the operational processes and delays,
average production rate, and current process variables like water flow, pressure and motor power.
By using a mathematical model of each motor/pump combination, the motor electrical power can be calculated when
the motor is equipped with a VSD and the speed of the pump is altered according to the planned new conditions
under which the water system will operate while still meeting the plant’s operational requirements. Similar
calculations can be done to quantify the effect of using a Soft Starter on a motor/pump combination.
From these analyses the expected average kW reduction over calendar time can be quantified after the energy
saving measures are implemented. The kWh energy savings and energy cost savings per annum can then be
determined.
Results – case study
Table 1 shows the potential average kW saving over calendar time under average conditions (e.g. average
production rate) for the steel plant under investigation. Note that these savings are only for the secondary water
systems that serve the HSM.
The production rate of the steel plant is, apart from other factors, influenced by operational delay duration. High
monthly delay durations will result in low production rates, and vice versa. At low monthly production rates, the
expected average kW saving is higher than at high monthly production rates. With the use of the energy savings
model for the steel plant the average kW saving performance curve as is shown in Fig. 8 was derived. Typically a
curve like this could be used in demand side contract contract to determine the monthly performance target that has
to be achieved.
http://www.ee.co.za/article/ist15406energyefficiencyimprovementinahotstripmillbyusingvariablespeeddrives.html 5/6
01/12/2015 Energy efficiency improvement in a hot strip mill by using variable speed drives EE Publishers
Table 1: Average kW saving for HSM water systems.
Secondary water system average kW saving
Fig.8: Average kW saving performance curve.
Conclusion
A hot strip mill in a modern steel plant presents energy saving opportunities, especially in the secondary water
systems that service the mills. In the case study it was found that an average saving of 3,290 kW over calendar
time is possible, which converts to R4,3m energy cost savings per annum at present electricity costs.
The project – for Mittal Steel Saldanha Steel – is currently in an approval phase at Eskom DSM and has still to be
implemented, therefore actual results are not available yet.
Acknowledgement
This paper was first presented at the Industrial and Commercial Use of Energy Conference organized by the Cape
Peninsula University of Technology in May 2006
Related Tags
energy efficiency (http://www.ee.co.za/tag/energyefficiency) hot strip mill (http://www.ee.co.za/tag/hotstripmill) IST
Otokon (http://www.ee.co.za/tag/istotokon) variable speed drives (http://www.ee.co.za/tag/variablespeeddrives)
http://www.ee.co.za/article/ist15406energyefficiencyimprovementinahotstripmillbyusingvariablespeeddrives.html 6/6
