Plant startup

Startup of the plant from a cold standby condition to a power producing condition requires the 

You can never have too many competent people on startup. 

The last, but very important, phase of a construction project is the plant startup. On a major project this is as significant an event as Front End Loading or Design Construction. 

On many major projects, startup is put on the back burner and not given the proper amount of planning and resources that are necessary to have a quick, successful demonstration of the operability. Although a well-planned startup will not overcome deficiencies in design and construction, a poorly organized one will seriously mar the completion of a well-designed and well-constructed plant. It is the responsibility of the Venture Manager to avoid this predicament. 

The successful startup must begin at the very onset of the project. Its planning must be part of the Project Execution Plan developed during Front End Loading  

Although this sounds like common sense, once the required resources for a startup are known, the most common management response is you don t need that many people. The Venture Manager must be firm in insisting on the required resources and keep reminding management of the old Irish saying  

---

United States. In 1980, Unocal began constmcting the Parachute Creek Project, designed to produce 1600 m (10,000 bbl) of upgraded shale oil per day. The project included a conventional underground room-and-pikar mine, the Unishale B (see Table 7) retort, and a special Unocal upgrading facihty. Plant startup occurred in 1986, and daily shale oil production reached 1100 m /d (7000 bbl/d). By 1991, total production exceeded 0.6 x 10 m (four million barrels). However, the Parachute Creek Project was shut down in mid-1991 for economic reasons. 

Approximately 97% of all turboexpander problems oeeur during the initial plant startup period. This eritieal period usually extends over several weeks, from the initial expander run until the plant pressures and temperatures are normalized and all related equipment is stabilized. 

Gans, M., Kiorpes, S. A., and Fitzgerald, F. A., Plant Startup—Step By Step By Step, Chemical Engineering, October 3, 1983, p. 99 (Inset Some afterthoughts on startups). 

More has been written about plant startup than one might imagine. Since considerable troubleshooting is needed during startup the two subjects tend to overlap. Therefore, the reader is encouraged to seek startup techniques also within the troubleshooting chapter. 

Anderson, G. D., Initial Controller Settings to Use at Plant Startup, Chemical Engineering, July 11, 1983, p. 113. 

Matley, J.. Keys to Successful Plant Startups, Chemical Engineering, September 8, 1969. p. 110. 

Troyan, J. E., series Hints for Plant Startup, Chemical Enginnering, Part I—Troubleshooting New Processes, November 14, 1960, p. 223 Part II—Troubleshooting New Equipment, March 20, 1961, p. 147 Part III—Pumps, Compressors, and Agitators, May 1, 1961, p. 91. 

Godard, K. E., Gas Plant Startup Problems. Hydrocarbon Processing, September 1973, p. 151. 

G. D. Anderson s article recommends initial controller settings for those control loops set on automatic rather than manual for a plant startup. For liquid level, the settings depend upon whether the sensor is a displacer type or differential pressure type, or a surge tank (or other surge) is installed in the process ... 

J. E. Troyan s series of articles on plant startup has a cause/effect table on instrumentation in Part II. This article also has troubleshooting hints for distillation, vacuum systems, heat transfer, and filtration. Here is the table on instrumentation. 

Boiler plant operational duties occasionally necessitate starting up boilers from cold (cold plant startups or dead plant startups), bringing a boiler online to support other operational boilers (live plant startups), and taking boilers offline. 

While startup and shutdown occur relatively infrequently in large continuous plants, they are inherent in batch plant operation. Most startup and shutdown procedures, whether devised empirically or theoretically, are designed to follow a recipe of actions with no feedback. Thus, if upsets occur, there is often no way to change the startup or shutdown in time to avoid imwanted process exclusions. Procedures are needed that incorporate feedback and adaptive techniques to the problem of plant startup and shutdown. 

The two indicators used in Example 10-5 give different answers. The Proceeds per Dollar of Outlay procedure is obviously wrong. It would take 12.5 years to pay off the investment for plant 1. If the 15,000,000 pretax profit available at the end of year 4 were invested at 7% interest, the accumulated interest would have exceeded 10,800,000 before plant 1 was paid for. In other words, the total profit for plant 1 would exceed 25,800,000 before plant 2 showed a profit. This is more than the total profit made by plant 2 during its whole life. If the 15,000,000 profit of plant 1 were invested for 21 years at 7% compounded annually, the total interest would amount to 47,100,000 or the total earnings would be 62,100,000. If after plant 2 were paid off die profits were invested at 7% per year, the interest would amount to 8,900,000 or a total pretax profit of 33900,000 25 years after the plant startup. This is 28,200,000 less than that earned in the same amount of time by plant 1. This shows that for this example the Annual Proceeds per Dollar of Outlay is a better economic indicator. 

The construction costs were incurred before the plant began operating and are spread unevenly over a number of years. Assume that on the average they date from a year before the plant startup. Time zero is assumed to be the beginning of the initial plant startup. 

An interest rate of 8% will be chosen. The fixed capital charges are assumed to have occurred a year before plant startup. The working capital costs occur during... 

Troyan, J.E. How to Prepare for Plant Startups in the Chemical Industries, Chemical Engineering, Sept. 5, 1960, p.107. 

Relief header isolation valve left closed during plant startup caused a heat exchanger to rupture. 

A second challenge was to rapidly develop real models as soon as possible after plant startup, in spite of the relatively small variation in composition expected at a given process point. This challenge was solved by modeling all three process points (esterifier exit, pipeline reactor exit, and pre-polymerizer exit) together. Inclusion of the process variability due to the startup itself also helped. The calibration space for the first calibration set is shown in Figure 11.7. 

New ideas for a process — Determining process chemistry —> Pilot plants studies — Plant design — Plant startup —> Debottlenecking or troubleshooting. 

Plant startup The commercial plant is started. In this phase, no actions for process intensification studies can be defined. 

In order to compare investments having different lives or with variations in return during their operating fives, it is necessary to convert rates of return to a common time basis for comparison. Although any time may be used for the comparison, the plant startup time is usually the most satisfactory. Expenditures prior to startup, and income and expenditures after startup, are converted to their worth at startup. The discussion that follows is based on the predicted startup time as the basis of calculation. 

---