Energy recovery projects often focus attention on fired heater duties. Burning less fuel presents an obvious benefit to a plant. Fuel costs money. The bill shows up easily as a direct charge from the fuel company. However, the basic rule in all energy projects is to track down the energy flows to something that you pay for. In plants, both simple and complex, this often may result in the age-old discussion of transfer prices versus overall site economics.
Consider the situation faced by a plant with a high-temperature reactor system. A fired process heater provided duty to get up to the required reaction temperature. A conventional feed/effluent exchanger system recovered duty upstream of the heater, reducing the heater duty. The feed/effluent heat recovery used five TEMA E-type shells in series.
The heater also included a steam generator system that provided medium-pressure steam for a variety of plant uses. Process demands set the heater duty requirement, causing steam generation to vary. The stack temperature out of the steam generator (convection) section stayed close to 520°F.
Temperature-cross pinches (www.ChemicalProcessing.com/articles/2007/070.html) in the feed/effluent exchangers limited heat recovery. One energy conservation project considered adding two feed/effluent exchangers to make a total of seven shells in series. At first glance, the energy recovery seemed to offer reasonable return on investment. Figure 1 shows a simplified plant sketch along with the heat balance boundary used for project analysis (solid boundary line). The new feed/effluent exchangers reduced process heater firing.
Has this analysis found the "right" answer? Further investigation raises serious doubts. Reduced firing on the process heater decreases the stack-gas flow rate to the convection section. The nearly constant stack temperature on the process heater results from a large convection section. The stack temperature pinches against steam production temperature. Therefore, the reduced heater firing, as set by the process requirement, also gives a nearly linear drop in steam production.
However, plant steam demand stays the same. So, some other part of the plant must provide the steam. In this case, the boiler firing goes up and boiler fuel demand rises. The added feed/effluent exchangers shift more duty to the boilers. The extended heat balance boundary (shown by the dashed line on Figure 1) gives the correct energy balance -- including the effect of the changes on the steam system.
What is the net energy savings of the project? A quick analysis compares stack temperatures. The boiler stack temperature runs around 350°F while the process heater stack temperature runs around 520°F. Yes, the project saves energy -- but not very much. Detailed analysis also would include efficiency differences between different fuel types, air preheat changes and many other factors. But by itself just comparing stack temperatures gives a rough 4%–4.5% energy savings on the extra duty provided by the two new exchangers. Few projects will make the cut with this low a benefit.
Here, the improper boundary resulted from not thinking through all the consequences of changes. However, one other issue continually arises: transfer prices. The current example uses cost of a stream (fuel gas) that the plant must purchase, allowing a relatively straightforward analysis because money changes hands. When money doesn't directly change hands, the evaluation can become tougher.
Complex plants may have operations run by many different business groups that are judged, and rewarded, by their specific results. Energy streams handed over from one group to another typically get transfer prices. Many different methods can be used to set these prices. Even in good faith, different pricing philosophies and logic can lead to transfer prices that result in strange conclusions.
For our example, imagine the steam generated by the process heater has a price set by its heat content. This ties its value to fuel. In contrast, imagine the price for the boiler steam has a fuel component plus an extra cost added for "guaranteed availability." The total boiler steam price will be much higher. So, the business unit responsible for running the process heater might want to reduce feed/effluent exchanger efficiency to generate more "low cost" steam. Opening bypasses will quickly lower efficiency. If the entire plant takes similar steps, the efficiency differences between the boilers and the process heaters add up. In extreme cases, boiler steam demand may drop so low that boiler operating problems occur.
Ideally, transfer prices should give the same answer as looking at an overall energy envelope. Sometimes they don't. When evaluating projects based on transfer prices always attempt to check the overall envelope as well. If the answers don't match, it's time for a thorough and careful look at the decision criteria.
ANDREW SLOLEY, Contributing Editor