• What is the optimum level of excess air in a natural draft heater?



  • rupam mukherjee, eil, rupammukherjee86@gmail.com

    I think it depends on number of factors. Is the furnace fired with fuel oil or it has fuel gas firing only? FO firing calls for higher excess air. Similarly, are you operating the furnace at design load or at turndown? Turndown conditions demand much higher excess air. Also, how reliable is your SIS or instrumentation system? A combustible analyser is certainly benefitial if it is intended to operate below the API recommended level. Also, one must verify the burner flame characteristics during the burner performance test at the vendor's workshop for low excess air levels. Trimming too much excess air and over dependence on instrumentation can lead to pitfalls. Better stick to API recommended excess air levels.


  • Richard Rhead, Refining & Aromatics Consulting, rick.rhead@gmail.com

    Others have provided good information on how to adjust and optimize. So I won't comment on that. But I will bring up some system limitations to be aware of as these might be preventing you from safely achieving the levels that you think you can reach.

    Fuel gas or fuel oil fired? Fuel oil needs more excess air than fuel gas. The second item to consider is fired heater construction: bolted, bolted and welded on one side, or bolted and welded on both sides? This is related to air leakage into the radiant section which is under vacuum. It is random leakage where you don't want it, instead of air at the burner tips.


  • Nitin Kirloskar, D.S.Kirloskar Consultants, nitinkirloskar@gmail.com

    While the experts have given good technical answers to the question, here I wish to bring up one important aspect of excess air control. To control excess air, one needs to 'measure' it correctly. There are good draft gauges available on the market, but often people tend to buy mechanical gauges, which won't offer you a control element. 

    Electronic draft gauges provide you with accurate readings as well as a control signal, which you can use to control excess air accurately. The good news is that these are available at almost the same price as mechanical gauges.



  • Jake Gotham, InSite Technical Services, jake.gotham@insitetechnical.com

    Previous answers have given various rules of thumb which are useful for design purposes or for initial screening to identify the most likely opportunities for improvements.  In a multi-burner natural draught furnace, it is very difficult to reliably set all the registers to give equal air to each burner, hence a certain amount of sub-optimisation is almost inevitable.

    To get maximum benefits from optimisation, and to ensure optimisation doesn’t create safety or environmental issues, the oxygen analyser should be coupled with a combustibles analyser.  Ideally the analysers should measure conditions below the convection section due to the large amount of air leakage in a typical convection section.

    The objective is to simultaneously minimise excess oxygen and combustibles.  One approach is to reduce air until the combustibles respond, then reverse the last movement on the air.  Another is to multiply the %O2 by the combustibles measurement, and minimise the product by adjusting the air.

    A skilled, experienced, committed operator is more important than the optimisation method, analyser type or control technology.  Edwin mentioned above the issue with operator priorities affecting their ability to commit time to continuous optimisation. I know of several sites that have tried to address this by creating an additional operator position dedicated to furnace tuning. This individual spends his shift moving from furnace to furnace optimising the firing conditions — obviously in close communication with the unit operator.  It is not difficult for this individual to cover the cost of his employment.


  • Grant Jacobson, Becht, gjacobson@becht.com

    It depends on the heater design and operating objectives. For most applications, optimised excess air (or excess oxygen if discussing on just an O2 basis) can be set between 15% and 25%. To confirm where this can be optimised safely it is important to conduct a CO breakthrough test and then set the optimisation targets based on where CO breakthrough is observed. An example of this would be if CO breakthrough occurs at 1.5% excess O2; the target to operate normally would be set at 3.0-4.0% excess O2. It is critical to not lose sight of keeping fired equipment stable and safe, and optimising to reasonable targets when able to do so.


  • Edwin Voeten, Petrogenium, Edwin.voeten@petrogenium.com

    Optimal value excess air should be evaluated on a case-by-case basis. However, as a design case 3 vol% O2 in flue gas is typically used, which corresponds to some 15% excess air over the stoichiometric requirement.

    The use of a forced draft fan is done for technical, economic, and/or historical reasons.

    The incentive to lower excess air is economics. As a guideline, some 1% fuel efficiency can be gained for every 2 vol% of O2 reduction in flue gas. Excess air is easier to control by use of a forced draft fan (together with other benefits from a forced draft arrangement). However, this requires additional capex. Consequentially, a forced draft arrangement is typically used in larger heaters.

    In a natural draft heater, for every process change or upset an operator may need to make changes to individual burner air registers in the field (and draft control/stack damper as well). Significantly, there may be an incentive to move towards forced draft where operation is expected to be less stable. Alternatively, in a natural draft unit a wider margin in excess air (a higher value than the design value) can be used to minimise the risk of sub-stoichiometric firing from any operational changes or upsets. This then has an economic penalty (see the numbers above).

    As a result, it is not uncommon to see a natural draft unit in the field operate in the 4-5 vol% O2 range, or higher. Although I should mention that this often is a consequence of operational staff having ‘other priorities’ and improvements are feasible.

    Note that this refers to gas fired natural draft units, which is the standard. I know of less than a handful of oil fired natural draft units (in Turkey, Kazakhstan, and Russia). Such units should not have been designed with the combination of fuel oil and natural draft, as this is a continuous source of trouble. Last thing I heard, the unit in Turkey was to be converted to forced draft.


  • John Skelland, KBC (A Yokogawa Company), John.Skelland@KBC.global

    Aim to operate the heater at its original design excess air level in the first instance. A good performance in modern heaters is 10% excess air (2% oxygen in the flue gas) for heaters fired with fuel gas. Fuel oil fired heaters usually require 15% excess air (about 3% stack oxygen). This is often due to fuel oil burners being slightly more difficult to tune and operate. Operating at these low levels is made easier if automatic control is installed rather than relying on frequent attention from the operators.

    It is sometimes possible to safely operate fired heaters at stack oxygen contents as low as 1% (~6% excess air) if an advanced control system such as Yokogawa’s Combustion One is installed. This system, incorporating tunable diode laser spectroscopy (TDLS), gives fully automatic control and optimisation of the combustion process. It measures oxygen, carbon monoxide, and methane concentrations in the flue gas to prevent the risk of unstable combustion.