CS-050: Fired Heater Design

CHALLENGE

The main challenge in this case study is the application of Flownex to:

• design the process by performing a heat and mass balance,
• size the fired heater and determine typical tube wall and flue gas temperatures,
• size the combustion air fan,
• determine the required ducting and piping sizes, and,
• analyze the combustion process, specifically the flue gas emissions.

BENEFITS

By creating a fired heater compound component, a Flownex model can be developed which is able to model both the fluid flow, combustion and heat transfer processes. This combined capability is not commonly available in other design tools.

SOLUTION

Flownex could effectively be used during the design and sizing of the fired heater, combustion air fan and the associated ducting.

INTRODUCTION

The design of a fired heater (or similar) package typically starts with a heat and mass balance. This first step is necessary to determine the process parameters which determine most of the sizing of the package. A fired heater package heat and mass balance includes the modelling of the combustion process and the modelling of the heat transfer and fluid flow processes. Modelling the combustion process typically involves the specification of the fuel gas composition, the combustion air composition, the air-fuel ratio and the fuel flow rate. The heat transfer from the fired heater combustion process into the process fluid may be specified in terms of an overall heater thermal efficiency. Typical heat and mass balance calculations do not provide estimations of the physical size of the heater or even the ducting and other components such as combustion air fans, neither do they enable the calculation of system pressure losses or heat losses. They are also incapable of providing insight into tube wall and process fluid film temperatures which are very important in the oil and gas industry. Flownex enables the user to perform all these tasks easily and quickly in a single calculation.

MODEL

The main purpose of this exercise is to size a fired heater (typically an API 560 design) and its associated combustion air fan and ducting. In order to perform this task efficiently, a Flownex model has been developed which is able to model the combustion, fluid flow and heat transfer processes. As such the model relies on a fired heater compound component and a few scripts to handle fuel and flue gas analyses. Required data inputs are indicated by blue rectangles in Figure 1. The fired heater compound component input and result fields are shown in Figure 2 and Figure 3 respectively.

Typically, the client would specify the following heater requirements:

• The required heater duty (to the process fluid).

• Process fluid composition and flow rate.

• Process fluid required inlet and outlet temperatures.

• A required, or at least a desired heater efficiency.

• Maximum allowable flue gas temperatures (if specified).

• Maximum allowable tube wall temperatures and/or process fluid film temperatures.

• Allowable process coil API erosional velocities, mass velocities or ρV2-values.

• Process coil fouling factors.

• Fuel gas composition.

• Ambient conditions.

The required heater data specification is as follows:

• Estimate the heater efficiency based on previous experience, or if the efficiency is stated by the client, specify this as the Required Efficiency (HHV). Typical fired heater HHV efficiencies range from 75% to 80%.

• Add estimates for the Heater Flue Pressure Drop and Process Pressure Drop. A typical flue side pressure drop is 50 kPa to 60 kPa. The model may be used to estimate the process side pressure drop by specifying the expected number of return bends in the coils and a return bend loss coefficient. These may be updated at a later stage when the actual heater pressure drop has been determined via dedicated heater design software.

• Specify the heater configuration; Counter Flow or Parallel Flow. Most fired heaters are counter flow. In parallel flow designs the process fluid inlet would be at the radiant section and the outlet at the top of the convection section. They are typically used in situations where corrosion may be a problem at the top-most convection tubes. This occurs when the top convection tube temperatures would be too close to the water or SOx acid gas dew point temperatures in counter flow configuration and therefore a risk of tube corrosion exists.

• Specify the Design Heat Flux. Adjust the Design Heat Flux to obtain larger or smaller coil areas. This parameter may also be used to obtain comfortable tube wall temperatures. This value may rely on previous experience and also depends on the type of process fluid. For fluids where local hot spots may cause degradation, such as oils, this value would be lower than others, such as air, where no such degradation occurs. 

  From practical experience, the following are approximate guidelines:

  • For water heating, a maximum of 36 kW/m2 should be used.

  • For oils with temperatures below 150°C, a maximum flux of 23 kW/m2 is recommended.

  • For oils with temperatures between 150°C and 250°C, a maximum flux of 21 kW/m2 is recommended.

  • For oils with temperatures above 250°C, a maximum flux of 19 kW/m2 is recommended.

  • For natural gases and other gases such as air the heat flux may be set higher, however a safe margin between the calculated tube wall temperature and the maximum design temperature for the project at hand should exist.

• Since this model does not perform detailed radiation or even convection calculations, but simply applies average heat fluxes over a coil area, peak fluxes and temperatures cannot be calculated. However, a Peak Heat Flux Factor is used to artificially increase the maximum temperature values to obtain an estimate of potential maximum tube wall temperatures. Therefore, a minimum value of 1 causes no peak temperatures. From experience, peak wall temperatures are about 5% to 10% higher than averaged maximum temperatures.

• If the heater is to feature both a radiant section as well as a convection section, such as API 560 Type C or Type E heaters, an estimate for the duty split between the radiant and convection sections is specified. This is achieved by specifying the Radiant Duty Fraction. This value is also used to control the Radiant Chamber Exit Temperature which cannot be too high, else the convection section will experience excessive temperatures, especially the tube fins. Typical bridge wall temperatures of 1000°C or lower should be obtained. From practical experience, radiant sections transfer between 50% and 60% of the heat. Therefore start the first iteration with a radiant duty fraction of 55%.

• The radiant section is sized based on the burner (or burners) flame size and layout. The flame dimensions are specified in terms of a Flame Length and a Flame Diameter for a single burner. Burner flame dimensions are available from the burner manufacturer for a specific duty rating. When multiple burners are used, the burners are arranged around a Burner Circle of which the Diameter must be specified. This diameter is zero for a single burner. Furthermore, a clearance must exist between the flame edge and the coil wall. These are typically 0.4m to 0.6m. These parameters will be used to determine the radiant section coil PCD.

• To estimate the required radiant section height, a Radiant Chamber Height to Width Ratio is specified. API 560 recommends a maximum value of 2.75 and a minimum of 1.5. For horizontally oriented heaters, API 560 recommends a range of H/W ratios for both maxima and minima depending on the heater duty. A ratio of 2.5 may be a good starting value. Another related factor specified by API 560 is a maximum volumetric heat release of 165 kW/m3 for gas fired heaters. Ensure that this limit is not exceeded with the selected geometry.

• The process fluid is selected as well as the coil materials.

• Coil inside and outside fouling factors may be specified. Fouling factors have a significant influence on tube wall temperatures.

• The coil geometry is specified in terms of the diameter, thickness and the number of parallel flow paths which essentially “duplicates” the coil into parallel flowing coils. Start by specifying a coil size and number of parallel flow paths that feels right from experience. After the first solution is obtained, the coil size and/or number of parallel flow paths may be adjusted depending on velocity and pressure drop considerations.

• Lastly, in the case of a convection section being present, a fin geometry may be specified. The model does not have the functionality to specify shock tubes - bare tubes without fins, usually the first 3 rows in the convection section. Since the model simply determines the required bare tube area from the specified Design Heat Flux, the specification of fins has no influence on the calculated required convection coil length. The only purpose of specifying the fin geometry is for the model to estimate likely fin tip temperatures and to calculate an equivalent average convection coefficient which only serves as a reality check. Typical convection coefficients range between 10 and 30 W/m2.K. According to API 560, fin lengths should not exceed 25.4mm and the minimum fin thickness is recommended to be 1.3mm. From practical experience, fin lengths may include 10mm, 12.5mm, 20mm and 25.4mm whilst the fin thickness may be either 1.3mm or 1.5mm. Fin densities should remain below approximately 200 fins/m according to API 560.

The heater, combustion air fan and ducting are then sized by following the steps below:

• Having specified a required or desired heater thermal efficiency, adjust the Fuel Gas Flow Rate to obtain the required Heat to Process (Duty) to achieve the required process fluid outlet temperature.

• Adjust the Excess Air % (Wet) to obtain the required Adiabatic Flame Temperature (theoretical maximum combustion temperature) and therefore indirectly the Flue Temperature into the convection section and into the stack. These temperatures depend on the combination of fuel flow rate, excess air percentage and the heater efficiency. Furthermore, the excess air also directly influences the O2 and CO concentrations in the flue gas. API 560 recommends 15% excess air for forced draft heaters and 20% excess air for natural draft heaters.

• Adjust the Radiant Duty Fraction if required, however this is a major unknown as the model does not really perform any radiation heat transfer calculations. Therefore aim to use the values suggested above.

• Adjust the coil geometry to satisfy velocity, mass velocity or ρV2-value requirements. Both the coil diameter and the number of parallel coils may be altered to achieve desired flow velocities. From practical experience, aim for velocities between 3 m/s and 4 m/s for liquids such as water. For more viscous liquids, the velocities may be significantly lower and the pressure drop may be the governing factor. For gases, aim for velocities of approximately 15 m/s.

• Adjust the Design Heat Flux. Again, when the process fluid is temperature sensitive, aim to use the values suggested above.

• Inspect the velocities in the air, fuel and flue piping and ducting and adjust their sizes to obtain comfortable velocities.

The combustion air fan flow rate is automatically set by the required combustion air flow rate calculated in the combustion calculations. The fan element then calculates the required duty point to satisfy the flow and pressure requirements.

SOLUTION

The model calculates the following:

• The combustion system performance. These include:
  • The calculated heat to process to achieve the required process fluid temperature increase;
  • Required burner heat release (HHV and LHV);
  • The heater thermal efficiency (HHV and LHV);
  • Combustion temperature;
  • Radiant chamber flue gas exit temperature;

The heater geometry as follows:

• The required coil lengths for both the radiant and convection sections;
• The required coil areas for both the radiant and convection sections;
• The radiant section chamber PCD and height;
• The flue gas flow results:
  • Flue gas mass flow and average actual volume flow;
  • Flue gas average density;
  • Flue gas side pressure drop;
• The process fluid results:
  • Process fluid mass flow;
  • Process fluid inlet and outlet temperatures for both the radiant and convection sections;
  • Process fluid inlet pressure and pressure drop;
  • Process fluid inlet and outlet velocities for both the radiant and convection coils;
  • API 14E erosional velocity, the mass velocity and the ρV2-value for both the radiant and convection coils
• The heat transfer results:
  • Heat transfer to each coil;
  • Average heat flux to each coil;
  • Radiant section volumetric heat release;
  • Convection section average convection coefficient;
  • Convection section estimated maximum fin tip temperature;
  • Coil inside and outside surface temperatures (minimum, maximum and average).
• the required fuel flow rate;
• the required air flow rate;
• The required combustion fan pressure rise and flow rate;
• Pressure drops for the various components;
• Duct and pipe velocities – check for duct and pipe sizing;
• Flue gas properties; of particular interest may be the oxygen content, the CO content, the SO2 content  and the dew point temperatures.

CASE STUDY

As an example, a natural gas fired heater package is to be designed to heat 73.6 kg/s of water flow from 155°C to 175°C. Perform a quick 15 minute heater package sizing and compare the results with a detailed final heater design done on a third-party fired-heater software such as HTRI or FRNC.

Data Specification

Typical fired heater properties when heating water are approximately as follows:

Table 1: Typical fired heater properties.

RESULTS COMPARISON & DISCUSSION

The following results were obtained using the Flownex model and are compared to results obtained from an actual detailed design done using HTRI:

Table 2: Results comparison.

Note 1: Due to the higher efficiency of the final design using HTRI, the HTRI calculated stack flue gas temperature is slightly lower. Changing the Flownex required efficiency to align with the HTRI results, a flue gas temperature within 10°C of the HTRI result is obtained.

Note 2: Note that the heater geometry was first determined using Flownex and the HTRI model subsequently used this geometry.

SUMMARY

Using Flownex, the design and sizing of the fired heater, combustion air fan and the associated piping and ducting could be performed within an hour. Furthermore, results such as process temperatures, tube wall temperatures, heat fluxes, flow velocities, pressure losses etc. could easily be obtained from the network. As an overall package sizing and design tool, Flownex has proven to be far more efficient than other detailed design software.