Starting Up a Furnace with Confidence: Part 1

Heine Weitz



Refractory failures are extremely costly, both from a safety and financial point of view. The capital cost for repairing or replacing the refractory lining and surrounding equipment is high but is usually overshadowed by the loss of production revenue.

The length of a furnace campaign is significantly affected by how well the furnace start-up was executed. We have seen a number of challenging projects in recent years. Most, if not all of these projects had refractory failures during or soon after starting up

It is worthwhile taking some time to define what we refer to when talking about a furnace “start-up”. Figure 1 shows the high-level structure of the phases executed on site during a furnace project. Although it doesn’t show a full breakdown of the phases, Figure 1 illustrates the typical phase sequence. One would also expect a slight overlap between some phases. Figure 1 also shows the stages in the hot commissioning phase. The highlighted area indicates the period that we refer to as furnace start-up, and the primary objective of this period is to get the furnace from a cold condition to steady-state without causing damage to the refractory materials or other equipment. The stages that form part of the furnace start-up are:

  • Heat-up: The furnace is heated by introducing hot gas into the furnace freeboard area, by passing electrical current through the electrodes, or by a combination of the two methods. The purpose of this stage is to gradually heat the refractory materials while avoiding thermal shock due to high stress gradients. A heat-up schedule is normally provided by the refractory supplier, which specifies the rate at which the temperature should be increased, as well as the duration of holding periods at critical temperatures.
  • Filling: During the filling stage, raw materials are fed through the feed system to build the alloy (or matte) and slag baths to their design levels.
  • Ramp-up: The furnace power is slowly ramped up during the filling stage, but is further increased during the ramp-up stage until design capacity is reached.

Figure 1: Phases in furnace project implementation.


Furnace start-ups are associated with significant uncertainty. Furnace design tends to focus on operation at full capacity since this is considered to be the most extreme operating conditions that the furnace will be subjected to. However, getting to design capacity can often be problematic, especially with a new furnace. Conditions change rapidly during start-up, and feedback from instruments is limited, which makes it challenging to understand what is going on inside the furnace.



Following are some key concepts that are important during furnace start-up:

  • Start-up burden: The material that is placed in the furnace prior to the heat-up. This material protects the hearth from thermal shock when switching on electrical power, and forms the initial bath. It normally consists of iron ingots, granulated iron, coke, or any other material suitable for the specific process.
  • Sacrificial lining: A thin layer of refractory material that is installed on top of the working lining. This is often a type of ramming material and provides protection against mechanical damage during furnace preparation, and chemical erosion during furnace filling.
  • Expansion allowance: Expansion papers are inserted between brick layers during refractory installation. This ensures that excessively high stresses are avoided by making sufficient allowance for refractory expansion during the start-up. These expansion papers burn away early in the heat-up.
  • Keying: The burning off of the expansion papers leaves gaps between the refractory bricks. Although the refractories will continue to expand during the ramp-up, we consider the refractories to have keyed when these gaps have been filled due to thermal expansion of the refractory bricks.


Furnace start-ups are fraught with risk and uncertainty. One of the most important is that of alloy penetration. Molten alloy introduced onto refractory bricks that have not yet keyed will penetrate the gaps between the bricks. Once this alloy solidifies, it will limit the extent to which the refractory bricks can expand, and cause excessively high stresses in the bricks. This poses a severe risk to hearth refractory integrity.

Refractory brick expansion is affected by the temperature ramp-up rate, length of the holding periods, start-up burden configuration, and sacrificial lining thickness, to name but a few. It is important to know at which point in the heat-up keying is expected to be complete for the specific configuration so that the filling stage is not started prematurely.

One also needs to understand how the installed thermocouple measurements relate to the freeboard temperature so that one can respond appropriately if the heat-up deviates from the plan.


The risk of damage and failure can be decreased substantially by doing the necessary work early in the project. Computational modelling is a comparatively inexpensive way of reducing uncertainty, and therefore risk, during furnace start-up. Modelling capabilities and computing power have moved forward in leaps and bounds, and we are able to accurately describe complex systems with computational models. Start-up data from clients where we have employed computational modelling confirm the success of this technique for describing a furnace during start-up.

In a follow-up post, we will discuss the aspects that we incorporate into such computational models to ensure that they are representative of the actual furnace. We will also look at some of the important questions that these models help us to answer. By taking due care with the modelling process, we equip ourselves with the necessary insight to reduce the risk substantially, help extend refractory life, and therefore provide the foundation for successful operation.

Recent Posts

Johan Zietsman


The format of the week has changed from last year to make the training more accessible and to equip attendees with the necessary knowledge and skill that will allow them to do powerful thermochemical calculations with FactSage™.


Dr Jaco Swanepoel 2024

Dr Jaco Swanepoel


Jaco is a Chemical Engineer with twelve years of experience in projects that range from reactive metals handling, reactive metal powder production in molten salt slurry reactors, extractive metallurgy to microwave-assisted gas-solid materials processing.
Jaco contributed to three process patents and obtained a PhD degree in Chemical Engineering. Jaco also completed the KTP Management and Leadership Study Program at Ashorne Hill Management College, England.

Stefan van der Merwe 2024

Stefan van der Merwe

Lead Engineer Digital

Stefan’s greatest passion is to enable people. By combining his greatest passion with his love for technology, he finds joy in making computing infrastructure accessible to the non-technical, and developing software that uncomplicates tasks. As a software engineer with qualifications in electronic engineering and bioengineering, Stefan is responsible for software development, enabling the team and clients to advance through insight.