Surviving a blackout.

Inside This Story

* Ford's Cleveland Casting plant was one of the many facilities affected by the famous blackout of 2003.

* Knowing that the time to power restoration was indefinite, the staff had to act quickly to ensure the facility's holding furnaces didn't suffer permanent damage.

* With strategic thinking in wake of the situation and efficiently utilizing resources, the facility was able to maintain and effectively restart its furnaces.

Imagine this: It is a blistering hot day in the middle of the summer, and your ferrous metalcasting facility is running nine channel induction furnaces at a time with 11 inductors on line. Everything appears to he in working condition, when all of the sudden--snap! The power's out. Your whole facility shuts down, rendering all casting and melting processes useless.

This is what happened to the Ford Motor Co. Cleveland Casting Plant, Cleveland. on Aug. 14, 2003, when much of the U.S. Northeast and parts of Canada suffered a mass power out age. Because the outage was so significant and widespread, there was no estimate to when the power would be restored, and the furnaces faced the risk of damage. However, by staying on the-ball during the blackout and the immediate hours after power restoration, the staff at the Cleveland Casting Plant and Duca Manufacturing Inc., Boardman, Ohio, personnel salvaged their holding line.

This article discusses how to condition and properly restart your facility's holding furnaces in the case of an unexpected power failure.

Think Fast

While the power outage did catch almost everyone involved off-guard, Cleveland Casting Plant's personnel knew they couldn't simply throw in the towel and ignore the situation until power was restored. The lack of an energy source forced the facility into circumstances that seemingly would harm any production, including cooling water and compressed air pressure dropping, cupola stack burdens cooling and inductor iron loops freezing. The staff knew that if inductors froze, they could be restarted if they were not damaged due to a lack of cooling water.

Because all of the furnace water systems switched over to city water for cooling of the critical bushings and cases, it appeared to be just a matter of time until the power was restored so that the furnaces could be restarted. However, six hours into the power outage, the situation deteriorated. Cleveland's city water supply was in jeopardy as the blackout had shut down all of the water pumping stations. Reportedly, there was only a two hour supply of city water left.

The loss of city water would mean a total loss of all of the furnaces unless something could be done to provide water cooling. Since the facility was equipped with a high-pressure fire loop water supply containing some 350,000 gallons of water, it was decided to connect each of the furnaces to this water supply. This involved a number of the staff members tapping the fire water lines, connecting hoses to those taps and running those hoses to the furnaces. The task took 5-6 hrs to complete.

Thirteen hours into the power outage, the city water was lost. Although most of the furnaces had been switched over to the high-pressure fire loop water supply before the city water stopped flowing, two furnaces were without water for 20-30 min. When water was restored to these furnaces, three hoses blew off due to steam formation. Once the hot water paths cooled, the hoses were reconnected.

Fifteen hours after the blackout began, the power came back on. Because the voltage was erratic, it was decided to wait until the voltage stabilized. When this occurred two hours later, each furnace was in turn switched from the fire loop water supply to recirculating water and was restarted. Four hours after the power returned, recirculating cooling water and power had been restored to all of the furnaces. The next task was to restore each furnace to its operating condition.

Seventy-two hours after the power initially went out, all nine furnaces were on line, up to temperature and capable for production.

Look at the Trends

While the methods used to bring the facility back to its working level might have been urgent, their effectiveness was based on trends of loop activity. Assume that before the power outage, the temperature in the loop was 2,650F (1,454C) and that the refractory in the inductor was at thermal equilibrium. As presented in Fig. 1, the rate at which the heat is transferred is dependent upon the temperature of the metal in the channel, the thermal conductivity of the refractory and the temperature of the water cooled case and bushing.

[FIGURE 1 OMITTED]

Consider what takes place at the 6 o'clock position in the loop (Fig. 2). Heat is conducted from the hotter loop and surrounding refractory to the colder case and bushing. The energy extracted causes the temperature of the loop iron and the surrounding refractory to fall. When the temperature of the loop falls below the temperature of the iron in the upper hearth, heat is conducted from the upper hearth to the loop. This heat slows the cooling process. It is for this reason that one should not decide to empty the furnaces.

[FIGURE 2 OMITTED]

Because solidification takes place first at the 6 o'clock position, the hotter metal atop the solidifying metal feeds the loop as a riser on a mold feeds the solidifying casting, which means that the loop is totally full upon solidification.

Rejuvenated Energy

To alleviate the holding furnaces from such harm, practical restart methods needed to be incorporated. The objective during the restarting process was to replace the energy lost during the power outage without doing any further damage to the furnace. When power is applied to a loop, the metal in the loop will be inductively heated while the refractory will not. This causes the loop to expand and stress the sintered refractory. How then can a frozen loop be heated without causing it to break, rendering tire inductor useless or to expand and crack the refractory lining resulting in a molten metal runout?

The technique involves repetitive heat and soak cycles. Power is first applied at a low level for a short period of time (known as the heat cycle). As the temperature of the loop increases, it expands and stresses the refractory. Power is then turned off for a longer period of time (known as the soak cycle). This causes heat to flow from the hotter loop to the colder refractory causing the loop temperature to decrease and contract the solid loop. As heat is conducted into the refractory, the temperature will in crease, and the refractory will expand. If the process is done correctly, the solid metal loop and the refractory will expand at virtually the same net rate. It is mandatory that this be done until the solidified metal begins to melt wherein the expansion forces begin to disappear. If the inductor is reheated faster than the controlled process, the sintered refractory will develop cracks, fun and run out.

The actual reheat cycle aider the blackout consisted of applying power at the lowest level for 1 min. and then turning it off for 9 min. This cycle was then repeated five times during the first hour. During each succeeding hour, the power-on time was increased by 1 min. while the power-off time was decreased by 1 min. After the 10th hour, there were no more heat and hold cycles as the loop is molten at this time. Power was then incrementally increased through the taps to full power to remelt and superheat the upper hearth. The important thing to remember was not to become concerned if the temperature of the molten iron in the upper hearth fell during the initial reheat period.

Nine furnaces were involved in the reheating process, and each furnace was started in the same manner. For example, a 65-ton pressure pour furnace had power applied at the lowest level of 200 kW. for 1 min., which imparted 3.33 kilowatt-hours (kWh) of energy into the loop. Since the heat content of cast iron in fine solid stage is 9.4 kWh/ton/100F, the loop temperature increased by about 142F (61C) during the first minute. As the temperature of the loop in creased, heat began to be conducted from the hotter loop to the colder refractory. This continued during the 9-min. power-off period so that at the end of the 10-min. cycle, the temperature of the loop and the refractory were virtually the same. This cycle was repeated five times during the first hour after which the power-on and power-off times were changed in accordance with the aforementioned schedule.

Once the temperature of the solid loop was raised to the melting point, which is at 2,066F (1,130C), the temperature of the loop didn't increase because the energy was being used to satisfy the heat of fusion requirements to melt tire iron. The heat of fusion to go from solid to liquid is 67 kWh/ton. Since the weight of the loop was about 500 lbs., a total of about 17 kwh was required to melt the loop. At a power level of 200 kW, the loop melted in 5-6 min.

Once the energy had been provided to melt the loop, further application of power superheated the molten metal in the loop. At 200 kW, 3.33 kwh of energy was imparted to tire loop each minute. Because the heat content of cast iron in the liquid stage was 10.4 kwh/ton/100F, the loop temperature should have increased by about 128F (53C) for each minute under power. However, because the molten iron in the upper portion of the loop was in contact with the colder upper hearth iron, heat was exchanged between the hotter loop iron and the colder upper hearth iron causing the upper case iron to melt and superheat, which limited the temperature rise of the loop iron.

Positive Results

To determine what effect there was to the inductor, resistance and reactance readings were compared (Table 1). During the first 14 days after the blackout, it appeared that the inductor exhibited some saturation and/ or wear based on a drop in the percent resistance and reactance, and after the power outage, the inductor exhibited buildup based on an increase in the percent resistance and reactance. Based on the data, it would appear that the power outage had little effect on the electrical characteristics of the inductor.

Because the inductors at the Cleve land Casting Plant had been fitted with type K thermocouples at the bushing, the status of the saturation network could be monitored. Knowing that cast iron solidifies at 2,066F (1,130C), the leading edge of the saturation network could be determined by measuring the temperature of the refractory at a given distance from the hushing. For example, if the bushing were at 140F (60C), and if the thermocouple that is placed at 1 in. from the bushing read 1,400F (760C), the saturation front would be 1.5 in. from the bushing (Fig. 3).

[FIGURE 1 OMITTED]

Once the furnaces were on line, the temperatures at the bushings were read in search of high temperatures, which would indicate excessive saturation or finning. Only two of the 11 inductors showed temperatures above 1,400F (760C). The temperatures at the bushing were monitored on an hourly basis for the first few days. In every case, once the furnaces were back to normal temperature, the temperatures at the bushing decreased.

The average depth of the unsaturated refractory was 2.65 in. before the power outage and 1.7 in. alter restarting the inductor. The increase in the saturated refractory depth didn't pose a problem because there was still adequate unsintered refractory depth.

Be Prepared

Cleveland Casting Plant's quick thinking to salvage its holding furnaces can set a baseline of what actions should be taken in lieu of a catastrophic event. Some precautions that should be considered in case of a blackout are having available diesel-powered contactor's air compressors and generators as well as additional water sources and, of course, flashlights and temporary lighting strings. Activity then should be prioritized from most important task to least, and roles should be assigned to the staff.

When such a power outage occurs, it's best to ensure that water is maintained on the bushing and inductor ease. Once the power is restored, it is mandatory to slowly reheat the metal that has solidified in the loop. Once the loop is remelted, the metal in the upper hearth can be remelted.

If the reheat procedure as presented is followed, a furnace can be returned to production with little or no damage. Ford's Cleveland Casting Plant only missed five shifts due to the outage, but if the emergency source of water wasn't used to cool the inductors, and if the furnaces weren't restarted properly, the power outage could have been much more devastating to the facility's production than what actually occurred.

Table 1. Electrical Characteristics of the Inductor

Date           07/21/03   08/04/03   08/16/03

Status          Startup    Before      After
Days                0        14          26
% Resistance      100        95         101
% Reactance       100        99         103

This article was adapted from a paper (04-159) presented at the 2004 Metalcasting Congress.

For more information

"Induction Furnace Electrical Fundamentals & Safety Circuits," D.S. Smalley, Proceedings of the AFS International Iron Melting Conference, Orlando, Fla., 2003.

About the Authors

Dour Rowe is a melt engineer for the Ford Motor Co. Cleveland Casting Plant, Cleveland. William.I. Duca is the president of Duca Manufacturing Inc., Boardman, Ohio.