Fig. 3 is a block flow diagram of this process al-ternative. The basic approach of Concept 2 is to provide the additional heat duty for reforming the larger process gas flow via the secondary reformer (SR). Using ambient air to supply the required larger oxygen flow to the reactor would introduce a considerable amount of excess nitrogen into the process gas. This can be avoided via operation with oxygen-enriched air.
The oxygen required for air enrichment must ei-ther be imported from outside or provided by an air separation plant within battery limits. In both cases, its operating and capital cost may not be neglected in the economic analysis. Generating the oxygen within battery limits is associated with significant additional capital investment. However, it has the advantage that the purity of the oxygen can be adjusted to the requirements of the process and in general does not have to be very high. Also, integration of the power re-quirements of the air separation with the plant's steam system is possible. Hence, this alternative has been assumed for the cost comparison.
The primary reformer remains mostly unchanged. Also, the other modifications required in the reforming section are fairly similar to Concept 1.
The basic approach of this alternative is an auto-thermal reformer in parallel to the existing re-forming section. Fig. 4 contains the block flow diagram of this process concept. In the same way as in Concept 2 it is operated by oxygen-enriched air in order not to exceed a reasonable amount of nitrogen in the process gas.
The autothermal reformer is a brick-lined vessel in which the two inlet streams of feed / steam mixture and oxygen are brought to reaction. It consists of a first reaction zone (or combustion zone) at the top, at the inlet of the two streams, and of a second catalyst-filled reaction zone in the bottom. Its principle design is shown in Fig. 5.
The existing reforming section remains essentially unchanged. Since the additional reforming is done entirely through autothermal reforming, this concept requires more oxygen than Concept 2. Thus the CO2 content in the process gas is even higher than in Concept 2, having an impact on the duty of the CO2 removal unit.
The three different reforming concepts of course have influence on the conditions and capacity of the other process units, but in principle for all concepts identical solutions for the other process units are selected for all three reforming options. The difference in equipment size has been taken into account for in the cost evaluation.
As the CO shift reactors are sufficiently sized in the reference plant, no change is made with them and a slightly higher CO content is tolerated. Also the methanation reactor does not require changes.
As there are many different processes available for CO2 removal, there are also many different options for a capacity increase. Options for ca-pacity increase include:
• Change of packing material in the absorber, allowing for higher gas and liquid loads,
• Installation of an additional flash step in the desorption section of the solvent cycle for improving the solvent regeneration,
• Complete change of solvent type (e.g. from potassium carbonate to amine-based), in-volving also significant modifications at equipment.
Typically, the absorber is the bottleneck. Changes in the desorption section are easier to implement as they involve low-pressure equip-ment. If needed, additional regeneration heat can be provided by low pressure stream.
Upgrading all items in the synthesis loop would be a task which would involve many modifications at the existing high-pressure piping and equipment. This would cause a high amount of modifications which would have to be executed in a small area in a short time while the plant is not in operation.
Therefore, an approach is selected where a whole new unit can be installed while the plant is in operation, and only a few tei-in points have to be connected during a shutdown of the plant. This is done by selecting the Uhde Dual-Pressure Synthesis [3]. It consists of a once-through (OT) ammonia synthesis which is added to the plant on an intermediate pressure level between synthesis gas generation and the synthesis loop as shown in Fig. 6.
For the envisaged relatively large capacity ex-pansion it can be assumed that the synthesis gas compressor is not able to cope with the signifi-cantly larger flowrate. Hence, an auxiliary com-pressor parallel to the first and second stage of the existing syngas compressor has been selected (not shown in Fig. 6), essentially taking the additional gas up to the intermediate pressure level.
The additional syngas is then mixed with the gas coming from the second stage of the existing syngas compressor and the combined gas is then passed through the OT synthesis. This process unit comprises of a gas / gas heat exchanger to provide the elevated converter inlet temperature, the actual OT synthesis converter, a steam gen-erator / boiler feed water preheater and a cooling train. The latter consists of a water cooler and a series of chillers, bringing the process gas temperature down to a level for separation of most of the generated ammonia by condensation.
Subsequently, the remaining process gas is passed on to the third stage of the synthesis gas compressor for further compression up to the pressure level of the ammonia synthesis loop. Essentially, the flowrate and composition of the gas fed to the synthesis loop is the same as in the original plant.
The Uhde Dual-Pressure Process has been suc-cessfully installed already for a revamp in a plant in Slovakia [4] and for two new plants [5], being the two largest single-train ammonia plants in the world.
One item particular to this plant is that the purge gas from the loop is not sent to a hydrogen recovery unit because it is already used for other purposes in the existing complex. This feature is maintained also for the revamp calculations. Since it is done like that for all investigated concepts, it does not affect the results of the study. Technically, there would be no difficulty to add a unit separating the hydrogen from the purge gas and returning it to the synthesis loop.
Steam System
For each process concept, the steam system is adjusted to match steam production from waste heat with steam consumption of process and turbines. Same as in the existing plant, also after the revamp, some MP steam has to be imported from outside battery limits.
Whole plant
Table 1 gives an overview which main equipment items have to be replaced or significantly modified for the three concepts. This list forms the basis for the capital cost assessment.
Results of the Economical Comparison
Operating Cost (OPEX)
The following streams entering or leaving the plant (see also Fig. 1) are associated with dedicated energy contents resp. cost data:
• Import: Feed gas, Fuel gas, MP steam, Electric energy.
• Export: Purge gas stream from ammonia synthesis.
Table 2 contains a representation of these streams for each revamp concept in terms of their energy content:
• Feed, fuel and purge gas export are repre-sented by their lower heating value (LHV).
• MP steam is represented by the fuel energy required for its production in a boiler.
• The imported electrical energy is shown as the fuel energy required for generating it via a steam cycle with an overall efficiency of 30 % (1 kWh electrical energy corresponds to 12000 kJ natural gas).
The overall consumption figures are shown in the bottom line of the table. Of course, the air separation unit with its motor-driven compressor is always included in the figures.
The consumption figures are relatively high compared to the values achieved by newly built ammonia plants. It is worth to mention that the study deals with an old plant and that the focus of the study is on capacity increase, not energy optimization. Certainly it would be possible to additionally improve the energy efficiency of the plant, but that would lead to additional cost. Since the changes would be about the same for all three concepts, they would not add value to the target of comparing the revamp concepts and thus are left out of consideration.
As the revamp parts are fully integrated into the existing plant it is not possible to give individual figures for the energy consumption of the revamp parts alone.
Capital Cost (CAPEX)
The capital cost for the revamp is estimated by first estimating the cost of the main equipment and then reflecting other costs for the imple-mentation by the factor method.
That means, first the equipment cost is deter-mined for the all the three process concepts using the process data determined in the simulations. Then the cost of all other contributions (piping, instrumentation, electrical, civil, engi-neering, procurement, erection and commissioning) is added, assuming that these can always be expressed by multiplication of the “pure” equipment cost by a certain factor. These factors are known from experience.
Concepts 2 and 3 require oxygen-enriched air for operation. It shall be noted that for the sake of a fair comparison the oxygen stream is not treated as a readily available utility but that the investment and operating cost of the air separation unit is included in the data.
Although the cost determined by the factor method includes cost for construction, it does not include the cost of lost production due to downtime for the revamp implementation. This is an important part of the real cost of a capacity expansion.
Obviously, erection time for the revamp extends over several months. During this time, the existing plant can maintain in operation. A complete shutdown of production is needed only to carry out the final tei-ins and for commissioning of the new sections.
However, an analysis of the revamp concepts reveals significant differences between them with respect to the activities for their final implementation. As the concepts vary mainly in the reforming section, the main differences are related to this equipment:
• Concept 3 requires only tie-ins at relatively cold and therefore non-critical piping.
• Concept 2 requires the installation and tie-in of a new secondary reformer. Hence, it is assumed that it demands at least one additional week for this work.
• Concept 1 requires difficult structural work to enlarge the box of the existing steam reformer. Even with a considerable amount of preassembling it seems likely that this work would prolong the scheduled shutdown by four weeks compared to Concept 3.
In the ammonia synthesis, the installation of the once-through section is proposed. This offers the same advantage of short time requirement for the tie-ins which Concept 3 offers for the re-forming as described above
The additional shutdown periods for Concepts 1 and 2 have been turned into capital costs via the assumption of an ammonia sales price of 400 USD/t and energy cost of 4.0 USD/MMBTU. This leads to the individual implementation cost for each revamp concept listed in Table 3.
Table 3 shows that Concept 1 is the one with the lowest erection cost. However, due to its complicated nature, the lengthy shutdown period adds significant cost by loss of production to it.
In overall cost, Concept 3 is the most attractive one. The difference between the overall costs of all concepts is 7 %. Certainly, there is some degree of inaccuracy in the cost data, but as the same methods of estimation were used for all concepts, it is believed that the data represent the correct ranking between the concepts.
CAPEX / OPEX Comparison
Finally, an economic comparison of all revamp concepts is made, considering their CAPEX and OPEX.
The comparison is made by determination of the specific production cost, that means the cost of production per ton of ammonia. (Alternatively, also the net present value of all concepts could be determined.)
This requires converting the investment cost into an annuity by an economic model, consisting of interest rate and required payback period. As it is always the case, the result of the comparison can strongly depend on the economic model used.
To illustrate the influence of the model, different scenarios have been evaluated. The first one (low interest rate and long payback period) in principle favors capital intensive plants with low specific energy consumption The second one (high interest rate and short payback period) just favors the opposite, i.e. plants with compara-tively low investment and higher energy con-sumption. The scenarios are combined with two different specific energy costs. The scenario parameters are listed in Table 4.
For operating cost, it is assumed that they are sufficiently covered by the contributions for the streams shown in Table 2. The assumption is justified that the other components related to e.g. personnel or maintenance are equal for all concepts. As the aim is only to determine the ranking between the concepts, they can be neglected. The study is made for two different gas cost as shown in Table 4. The cost for steam and electricity is derived from the gas price.
The resulting specific production cost (CAPEX plus OPEX) are summarized for different com-binations of parameters in Table 5. It shall be mentioned again that these figures shall serve only to determine the ranking between the con-cepts. As the energy consumption of the plant used as the basis is not optimized, some figures appear high.
The ATR-based revamp Concept 3 with the lowest specific energy consumption (see Table 2) is also the one with the lowest investment cost. Hence it is not surprising that it also shows the lowest overall production costs.
Carbondioxid Production
For all concepts, CO2 is emitted by the following three sources:
• flue gas from reformer stack (ISBL)
• flue gas from steam generation boiler stack (OSBL)
• CO2 stream from CO2 removal unit. Practically all carbon in the natural gas is finally ending up as CO2 in one of the above streams.
Also to the electricity consumption a (virtual) CO2 emission can be assigned because electricity production (inside or outside the plant) is linked to CO2 emission. As the electricity consumption of all three process concepts is similar and fairly small compared to the natural gas consumption, for sake of simplicity, no CO2 emission equiva-lent is assigned to electricity consumption. Table 6 shows the thus determined CO2 emission per ton of ammonia.
While the first two of the above listed streams from the stacks contain CO2 in a concentration of only approx. 10 % at ambient pressure, the third one is more than 99.5 % pure CO2 at slightly elevated pressure. Only this CO2 can readily be used for production of urea fertilizer.
Considering the stoichiometry of urea formation from ammonia and CO2 and the available amounts of the reactants, Concepts 2 and 3 show an increase of about 6 % in maximum possible urea formation compared to Concept 1 (see Table 6).
If one would like to achieve the same urea pro-duction by the other two concepts, their amount of usable CO2 would have to be raised. Basically two options are available to do so:
• CO2 can be washed out of the flue gas of re-former or boiler by absorption and desorption. The result is an almost pure CO2 stream which could be added to the existing CO2 stream. This solution adds high investment cost and a little operating cost to the process.
• More gas can be fed through the plant up to and including the CO2 removal unit. This would produce more CO2 in the process which is consequently separated in the CO2 removal unit, thus increasing the CO2 export stream. The surplus synthesis gas downstream of the CO2removal is fed to the reformer burners. This solution makes the reforming and CO2 removal sections of the plant a little larger, increasing investment a little, but adds significant operating cost to them. Both concepts have been applied already to re-vamps and new plants.
Summary
A detailed investigation has been carried out to assess the economic feasibility of three different concepts for a capacity increase by 30 % of an existing ammonia plant. The main difference be-tween the three concepts is in the reforming section.
The aim of the investigation is to establish an economic ranking between the revamp concepts. Concept 3 which is based on an autothermal reformer (ATR) turns out to be the most attractive solution. The other concepts (enlargement of the existing steam reformer resp. operation of the secondary reformer with oxygen-enriched air) have higher overall costs.
Responsible for this ranking are mainly the advantages of the ATR-based concept in overall capital costs compared with the other concepts. The process calculations show only moderate differences between the individual energy con-sumption figures.
The overall capital costs must include all costs associated with revamp implementation, including loss in production by longer plant shutdown. Especially Concept 1 (enlargement of the existing steam reformer) requires considerably more complicated implementation work which con-tributes to its higher overall cost.
References
[1] M. Papsch: Technische und wirtschaftliche Bewertung von Prozesskonzepten zur Synthe-segaserzeugung in Ammoniak-Anlagen, Bachelor Thesis, University of Applied Sciences, Krefeld (2011)
[2] J. Johanning, K. Noelker: Comparison of syn-thesis gas generation concepts for capacity en-largements of ammonia plants, Nitrogen + Syn-gas 2006 International Conference, Athens (2012)
[3] J. S. Larsen, D. Lippmann: The Uhde Dual Pressure Process – Reliability Issues and Scale Up Considerations, 47th Annual Safety in Am-monia Plants and Related Facilities Symposium,
San Diego, California (2002)
[4] F. Kessler et al., First application of Uhde’s dual pressure ammonia process for revamping of the Duslo ammonia plant, Nitrogen + Syngas In-ternational Conference 2006, Vienna (2006)
[5] K. Noelker: Commissioning Experience of the World’s Largest Ammonia Plant, ACHEMA International Conference 2009, Frankfurt a.M. (2009)
* K. Noelker ist Head of Process Department der ThyssenKrupp Uhde GmbH, Dortmund, Deutschland. Copyright 2012, American Institute of Chemical Engineers, Volume 53, "Safety in Ammonia Plants and Related Facilities," pp. 281-294. Reprinted with permission. The 2013 Ammonia Plant Safety Conference will be held at the Marriott Hotel in Frankfurt-am-Main, Germany, from Aug. 26-29
No comments:
Post a Comment