Desktop version

Home arrow Engineering

  • Increase font
  • Decrease font


<<   CONTENTS   >>

LCA approach

2.3.1 Sources of CO, in a natural gas processing

To ensure that the comparison with existing technologies is fair, all major CO, emissions across different pathways should be considered. In particular, the LCA calculations include the following sources of CO, in the processes.

2.3.1.1 Emissions in natural gas upstream plants

Production of natural gas at the field and its subsequent transportation to the downstream contribute significantly to the CO, emissions. In the present work, the GREET (Greenhouse Regulated Emissions and Energy use in Transportation) model was used. This software was developed by Argonne National Laboratory, and is considered to be a reliable benchmark calculator for taking into account NOx. CH4 and CO, from the pertinent source of emission under consideration. The NOx and CH4 emissions are calculated as equivalents of CO, emissions by using a multiplication factor, which is a function of the relative greenhouse gas potential of CH4 and NOx in comparison to CO,. More details on this are given in Afzal et al. (Afzal et al., 2018). For all calculations, a typical natural gas composition of 95% methane with a molecular weight of 16.81 kg/kmol was assumed. After incorporation of these values, the total CO, equivalent emission in the upstream is estimated to be about 620 g/kg of natural gas delivered at the GTL plant. This number, in other words, is an equivalent of 0.25 mole CO, produced per mole of methane processed.

2.3.1.2 Energy requirements in the reformer

An approach to calculate the energy requirements in the reformer is provided in section 2.2 of this chapter. All the conventional pathways to produce syngas (SRM, POX and ATR) lead to indirect CO, emissions resulting from heat duty for operating these units. For instance, in SRM, the contribution is both from the endothermic heat to boil the water at room temperature for production of steam of sufficient quality to feed the reformer in addition to the endothermic heat required to drive the reforming reaction. As for POX, even though it is a net energy producer, a huge energy footprint is associated with the preheater that heats the feed to approximately 1000 °C before it is allowed to enter the reactor. A DRM reaction, on the other hand, does not require a boiler unit like SRM. nor a pre-lieater like POX for high temperature feed, but it requires extremely high heat duty via a fired heater to drive the highly endothermic reaction. Due to the large capacity and high-energy demands, as stated above, the reformer block in a GTL plant is one of the most energy intensive processes, and accounts for at least 40% of total plant energy. As per the literature available for industrial standards, 85% efficiency of the reformer block was used for all the calculations. In addition to this, a heat credit for the exhaust gases leaving at high temperatures from the reformer reactor was also included, as they may be used to heat another unit or produce high quality steam for power generation.

2.3.1.3 Oxidant production

The oxidants used in the SRM and POX reforming reactions are steam and oxygen, respectively. The role of an oxidant is to provide oxygen for the conversion of methane to carbon monoxide and hydrogen. There are carbon footprints associated with the production of steam in a steam boiler and oxygen in an air separation unit, therefore, they have to be accounted for in the overall LCA of the pertinent reaction. The main contributor to the carbon footprint of CO, in the DRM process is the energy required for its separation from flue gases and other major sources of CO, in the plant. Table 1 below pror ides carbon footprints for SRM, POX and DRM processes.

Table 1. Footprints associated with oxidant production.

Oxidant

Carbon footprint in terms of СОг

Steam

471 [g/kg steam]

Oxygen

273 [g/kg oxygen]

CO,

63 [g/kg CO,]

The data provided in Table 1 for steam and oxygen is from GREET model, while for CO„ it is adopted from a separate study done by David and Herzog (David and Herzog, 2000). Assuming a 3% concentration of CO, in flue gas and 90% capture efficiency, the study by David and Herzog reported a CO, equivalent emission of 63 g/kg of CO,. The prediction from the GREET model for O, and H,0, however, only accounts for the energy-related CO, emission resulting from the production of pure O, and H,0. The approach implies a very conservative estimation, as a very dilute CO, concentration of flue gas is considered. However, a concentrated flue gas with a higher amount of CO, can safely enable exclusion of separation related emission from the overall calculation due to its negligible quantity.

2.3.1.4 Catalyst regeneration

Although the industrial conditions of the reformer are optimized to operate convincingly under the carbon fr ee zone, the geometrical considerations of the reactor tube could still result in carbon formation. A typical industrial reformer tube is up to 13 m long and 0.1 m wide in a cylindrical form. The tubes are designed in such a way that maximum contact time is provided in a tortuous path that enables the highest utilization of the packed catalyst. In addition to this, the tubes operate at 20 bar pressure to overcome pressure drop, and to meet target production rates. Due to these conditions, there is a possibility of coke formation in localized regions of the reactor tube. This could not only disrupt the favorable coke-fr ee equilibrium condition, but also lead to slow progression of deactivation throughout the reactor bed. Due to this reason, the catalyst bed needs to be regenerated frequently in order to avoid consequent issues of pressure drop and deteriorating syngas quality owing to coke formation. A typical procedure to regenerate the catalyst bed involves oxidation with air or pure oxygen to form carbon dioxide, which leads to CO, emissions. This has been incorporated in the calculations where 1 mole of carbon produces 1 mole of CO, during regeneration.

2.3.1.5 CO, from reforming reactions

Some of the side reactions in reforming processes produce CO,. Since these reactions are coupled with primary reactions, they cannot be completely eliminated. However, the selectivity of these reactions could be reduced by using efficient catalyst materials and with the selection of proper operational conditions. The CO, forming reactions in particular are the Boudouard reaction (2CO —> C + CO,) and the water-gas shift reaction (CO + H,0 —> CO, + H,). These reactions contribute directly to the CO, emission as CO, is a part of the products. The LCA model accounts for such reactions by applying negative credits for these reactions in overall CO, emission calculation.

2.3.1.6 Optimization model formulation

The target for optimization in the present study is to maximize the number of moles of syngas produced while maintaining a fixed syngas ratio. The approach used in this work is adopted from a detailed methodology illustrated in El-Halwagi et al. (El-Halwagi et al., 2017). A systematic procedure adopted from the original work of El-Halwagi et al. (El-Halwagi et al., 2017) and Afzal et al. (Afzal et al., 2018) is presented below in context of the subject of reforming processes:

Objective function:

Maximize syngas production.

Constraints:

• Carbon footprint reduction <

where, is increased iteratively from target for minimum carbon footprint

  • • Carbon footprint calculation = E(Carbon footprint in terms of CO, equivalents)
  • • Reforming equilibrium calculation using Gibbs free energy minimization
  • • H,:CO ratio to be changed iteratively from a low value of 1 to a high value of 4 so as to assess all the possible case scenarios and combinations of reforming processes

The equations for calculation of CO, emissions of each reforming process provided in section 2.3.1 are given below:

Carbon footprint of SRM=

{[CO, equivalent in reformer outlet]

+ [CO, equivalent from SRM furnace duty supplied by Natural Gas]

+ [CO, equivalent of upstream emissions of Natural gas and feed]

+ [CO, equivalent due to steam production]

+ [CO, equivalent of catalyst regeneration-burning of coke]

- [CO, equivalent credit due to reformer outlet heat integration and reduction in upstream emissions due to reduced use of natural gas]}

Carbon footprint of POX=

{[CO, equivalent in reformer outlet]

+ [CO, equivalent from POX furnace duty supplied by Natural Gas]

+ [CO, equivalent of upstream emissions of Natural gas and feed]

+ [CO, equivalent due to oxygen production]

+ [CO, equivalent of catalyst regeneration-burning of coke]

- [CO, equivalent credit due to reformer outlet heat integration and reduction in upstream emissions due to reduced use of natural gas]}

Carbon footprint of DRM=

{[CO, equivalent in reformer outlet]

+ [CO, equivalent from DRM furnace duty supplied by Natural Gas]

+ [CO, equivalent of upstream emissions of Natural gas and feed]

+ [CO, equivalent due to CO, separation]

+ [CO, equivalent of catalyst regeneration-burning of coke]

  • - [CO, equivalent credit due to reformer outlet heat integration and reduction in upstream emissions due to reduced use of natural gas]
  • - [CO, in reformer feed]}

The last term in the carbon footprint equation for DRM represents carbon fixation achieved by DRM. If the value of this quantity is greater than the total CO, footprint, then a net CO, fixation is realized and DRM is a CO, sink. Otherwise, the process is still a net CO, producer.

Additional constraints applied to the reformer in lieu of fan comparison between the three reforming processes are provided below: о х. ^inOUt.H;

• Syngas ratio = —-

nout.CO

• Operational temperatures of the reformers:

SRM: 850 °C to 1000 °C

POX: 1300 °C to 1400 °C ATR: 950 °C to 1100 °C

• Total methane fed to each reactor:

In rH = 1

Ш.СН4

• Oxygen flow to POX or ATR reactor:

пш.о,: 0 t0 2

• CO, flow for DRM:

■Vcoy 0 t0 2

• Steam to Carbon (S/C) ratio: --

Пт.СН4 +nm.CO,

• For SRlvI,

^in.H,0* 2 t0 4

  • • For POX,
  • •Vh,o: 01 x Пш.СН4
  • • For DRM,

nm.H,0: (nm.CH4+ Пш.СО,)

The reported optimization algorithm was implemented in LINGO* and solutions were computed for various case scenarios pertaining to a syngas ratio ranging from 1 to 4.

 
<<   CONTENTS   >>

Related topics