CC USE COLUMN DATA TO INFER AND CONTROL CRUDE FRACTIONATION PRODUCT PROPERTIES

Transcription

1 NATIONAL PETROCHEMICAL & REFINERS ASSOCIATION 1899 L STREET, N.W., SUITE 1000 WASHINGTON, D.C USE COLUMN DATA TO INFER AND CONTROL CRUDE FRACTIONATION PRODUCT PROPERTIES By Mark Schuler Control Systems Engineer United Refining Company Warren, Pennsylvania Y. Zak Friedman Principal Petrocontrol New York, New York And Michael G. Kesler President Paul Belanger Kesler Engineering, Inc. Brunswick, New Jersey Presented at the NPRA 2000 Computer Conference November 13-15, 2000 Palmer House Hilton Hotel Chicago, Illinois

2 This paper has been reproduced for the author or authors as a courtesy by the National Petrochemical & Refiners Association. Publication of this paper does not signify that the contents necessarily reflect the opinions of the NPRA, its officers, directors, members, or staff. NPRA claims no copyright in this work. Requests for authorization to quote or use the contents should be addressed directly to the author(s)

3 Abstract Use Column Data To Infer and Control Crude Fractionation Product Properties Mark Schuler United Refining Corporation Y. Zak Friedman Petrocontrol Michael G. Kesler & Paul Belanger - Kesler Engineering, Inc. Qualities of products of crude fractionation tend to vary because of changing crudes and operating conditions. On-line measurement of these properties is costly and, often, unreliable off-line, lab measurements are not timely enough to help operators keep the qualities on target. GCC (Generalized Cutpoint Calculation) is an inferential model for estimating crude and product properties. It uses heat balances, standard conversions of EFV to TBP values, and semi-rigorous engineering techniques to construct the crude TBP from column data. It, then, predicts, and calculates set-points for product flows to maintain on-target product qualities. It also calculates set-points of pumparound flows to maintain satisfactory internal-reflux flows. It focuses on heat balances for fast detection of heat disturbances due to changes in crude composition. As a result, GCC is able to predict and control product properties through major crude switches, as has been demonstrated in the several dozen crude unit installations of GCC. To increase speed and simplicity, GCC takes into account a finite set of instrument measurements. However, with no redundancy GCC could lose accuracy, should one of these measurements be erroneous. CPPM (Crude Plant Performance Monitor) is a rigorous distillation program, used to analyze all data available around the crude column: measurements of temperatures, pressures, and product flows. CPPM works by reconciling the crude distillation profile vs. vol% (TBP) against process measurements. CPPM uses redundant measurements to construct the crude TBP, and should a measurement be erroneous CPPM can detect and disregard the measurement. Once reconciliation is achieved, CPPM calculates product ASTM s, V/L profiles, and other column parameters. The United Refining Corporation (URC) system incorporates GCC and CPPM working side by side. GCC works every minute, at all times and particularly through crude switches. CPPM works every several minutes and whenever it detects steady state. It issues a set of biases to correct the GCC inferences. Combining the two models retained the speed and robustness of GCC and the rigor of CPPM. The paper presents results that show dramatic improvements in maintaining on-target product qualities compared with data preceding system installation. The paper also discusses also the system architecture and its integration with PI and ProcessBook, products of OSI Software, Inc. Page 1

4 BACKGROUND The paper describes recent installation of an on-line closed-loop system to control product qualities of the Preflash and Atmospheric Columns in a 65,000 BPD refinery of United Refining Corporation (URC) in Warren, PA. Figure 1 is a schematic of the two columns. The crude, after preheat, enters the Preflash Column at about 320 F. An additional small feed, Wild Naphtha of varying composition, enters the top tray of the Preflash Column with the reflux. The reflux controls the top-tray temperature, and hence the heavy-tail composition of the Light Naphtha product. A small off-gas stream leaves the Overhead Receiver. The bottoms of the Preflash Column flow to the Atmospheric Heater. Here it is heated to about 700 F before entering the Flash Zone of the Atmospheric Column. The top product is Heavy Naphtha, whose heavy- tail composition is controlled by reflux and top-pumparound. Three side products (Kero/Jet, Furnace Oil (F.O.), and Wax Oil (W.O.)) are stripped with steam in the side strippers. The Reduced Crude goes directly to the Vacuum Heater. Two pumparounds control heat removal and internal reflux. Operations of the column present several noteworthy problems: The crude flow and composition vary significantly, as shown in Figure 2. Major switches of crude occur almost daily. Crude is often not uniformly blended. Lab analyses of products are typically done every four hours. Varying crude composition makes it difficult for the operator to interpret past lab results for current conditions, particularly if a sample was pulled just prior to, or during a crude switch. The composition and flow of the Wild Naphtha is variable and not well-known, compromising control of the Light Naphtha tail end. Draw-off trays in the Kero and Furnace Oil sections of the Atmospheric Column often run dry, due to tray leakage/weeping. Measuring instruments of the Atmospheric Column Reflux and the Upper and Middle Circulating Refluxes often bump up against process constraints. As a result of these difficulties, the quality of products varies widely from desired specifications. Figures 3, 4 & 5 show typical 90% ASTM s of Heavy Naphtha, Kero, and Fuel Oil, during one month s operation this year, prior to installing the subject system. The temperature spread in these figures often exceeds 50 F, well outside the desired range of product qualities. Page 2

5 PROJECT OVERVIEW The main objective of this project has been to improve product-quality control, and in particular, to control the columns and product qualities during crude switch-overs. Another objective has been to improve control of pumparound heat removal, hence internal reflux of the Atmospheric Column. To achieve these objectives, the project utilized inferential technology, which included: i. Use of temperature profiles, pressures, product flows and other process measurements of the two columns, to construct the TBP of the crude being fed to the unit. ii. Prediction of product qualities and V/L profiles, based on the constructed TBP of the crude. iii. Calculation of set-points of side-stream and pumparound flows, to achieve desired product qualities and reasonable internal reflux. Two main programs, working side-by-side, have been installed in the plant, to implement the inferential technology. GCC the Generalized Cut Point Controller, a product of PetroControl, and CPPM the Crude Plant Performance Monitor, a product of Kesler Engineering, Inc. (KEI). GCC is a compact, integrated package of constraints, dynamics, manipulated variables and inferential models. It uses assumptions that enhance its speed and robustness (Reference 1). CPPM uses, at its core, a rigorous, proprietary tray-to-tray model (Reference 2) and redundant measurements that enhance its accuracy. It periodically updates selected parameters to correct the GCC biases (See Figure 6). Combining CPPM s rigorous solution with the speed-ofresponse and robustness of GCC particularly during crude switchovers has resulted in superior prediction and control of inferred product ASTM s, e.g., of Furnace Oil. The two programs, GCC and CPPM, are integrated with PI, a product of OSI Software, Inc. to retrieve on-line data and store results in real time. Set-point values that GCC stores in PI are relayed, via an interface, to DCS for closed-loop control of selected manipulated variables. This PI is a dedicated, local system, not connected to the refinery information system for reasons of Page 3

6 security. In addition to the control functions, PI is used for performance monitoring of the application. Further details of GCC, CPPM and the overall system architecture follow. GCC GCC performs the following functions: Constructs a crude TBP Predicts product properties and V/L profiles Calculates set-points for product flows required to meet quality specs and for pumparound flows to provide reasonable internal reflux i. Constructing the Crude TBP. GCC calculates the portion of the crude vaporized, at two anchor points: the column flash-zone and top-tray overhead. It assumes a linear slope for the crude TBP between the two anchor points. To calculate the TBP of the crude at the flashzone, GCC considers the temperature of the flash-zone, corrected for partial pressure, to be EFV (Equilibrium Flash Vaporization) end-point (dew point) of the vaporized portion of the crude. It uses API data book procedures (Reference 3) to convert the EFV to TBP values. Further, GCC uses a heat balance to calculate the portion of the crude vaporized, rather than obtaining it from the sum of the products. This approach is significantly more accurate, particularly during unsteady operations following crude switchovers. One reason is that a shock resulting from changes in mass flow e.g., due to a lighter crude can be masked by tray and vessel hold-ups. A second reason is that changes in heat- flux can not be hidden and are immediately detected by a heat balance. GCC calculates the TBP of the lower-boiling anchor of the crude, in a similar fashion, by considering the top-tray temperature, corrected for partial pressure, to be the EFV of the endpoint of that vaporized portion of the crude, namely, the Heavy Naphtha. GCC then constructs the TBP curve of the crude. Generally, GCC uses the sidestream draw temperatures to determine the degree of nonlinearity in the curve. However, for this project, Page 4

7 given the better steady state results of CPPM, GCC assumes the TBP profile of the crude between the two anchor points to be linear. ii. Prediction of product ASTM s. Having established a TBP profile, GCC calculates product flows from heat-balances around each side-stream section, and product cut- points from the crude TBP vs. vol%. It uses API procedures to convert the TBP s to ASTM s, and short-cut methods to correct tail-end ASTM s for internal reflux. iii. Calculating Set-Points. GCC provides on-time set-point values of key manipulated variables to DCS via PI. These values insure on-spec product yields and reasonable internal V/L profiles. GCC performs the above functions approximately once every minute. Its speed - as well as robustness is due, in part, to the short-cut methods that it uses. GCC calibrates the correlations and the model that it generates with several sets of historic data. CPPM, described in the next paragraphs, provides a more rigorous basis for tuning key parameters of GCC, approximately once every 10 minutes. An important aspect of the on-line implementation is to insure validity and reasonableness of the plant data obtained from PI. GCC includes data conditioning procedures to check the data, and to identify outliers, faulty instrumentation, etc. Figure 6a shows GCC controls in the plant environment. CPPM The core of CPPM is a rigorous tray-to-tray model of the two columns. The T-t-T calculations use open-equation formulation with proprietary procedures of lumping components/trays to enhance solution speed and robustness (Reference 2). The model has been calibrated with several historic data sets, to establish tray efficiencies in the various sections of the columns. The calibrated model has been incorporated in a solver/optimizer, which retrieves from PI a product of OSI Software, Inc. temperature profiles and product flows, and constructs interactively a crude TBP, by minimizing errors between calculated and measured values. Page 5

8 Having established the composition of the crude, CPPM predicts, with the T-t-T model, 90% - ASTM s of the key products, as well as V/L profiles for the given set of pumparound duties. It stores the calculated values in PI. CPPM also calculates and stores in PI updated parameters for GCC, approximately every 10 minutes. Figure 7 presents an overview of the system. The figure shows two blocks essential for on-line, closed loop operations; namely, Data Conditioning and, Checking Results. CPPM includes statistical tests and heuristics to eliminate inconsistent or suspect data. It also checks to insure the column has reached a reasonable steady state of operations. Similarly, CPPM checks reasonableness and consistency of results, limiting changes to a certain percentage of proceeding values. These filters have improved performance and conformity of the predicted qualities with lab measurements. RESULTS AND CONCLUSIONS Figure 8 shows an example of dramatically improved control of Heavy Naphtha 90% ASTM at closed-loop, during a 5-hr period, compared with open-loop operation. Figure 9 shows 90% ASTM closed-loop control of Kero (lower graph) and Furnace Oil vs. lab data. As can be seen, the system predicts values that match well with the lab measurements. Furthermore, it is obvious that the current product 90%-ASTM s are significantly closer to desired specs, and have a much narrower scatter, than those recorded before installing the system, shown in Figures 3-5. These results show significant benefits beyond original expectations. Other benefits of the project have been: Shortening crude switchover time from about four hours to 90 minutes, and maintaining product qualities during crude switchovers. Better control of internal reflux, reducing the frequency of side-stripper level loss events from daily occurrences to once every two months. Higher unit efficiency and increased throughput due to better internal-reflux control. Reduction of operator loads. In terms of operator acceptance, following training and a period of open loop operations, closing Page 6

9 the quality-control loop was accomplished with relatively little trouble. Then the application quickly gained popularity, due to its ability to handle crude switches effectively. Several months later the internal reflux control part was tuned, reducing the frequency of sidestream level losses. Since then (8/15/00), the application has run practically all the time. Recently the operators noticed that the PI displays provide more and better information than DCS. They requested and will shortly receive a PI screen in the control room, showing trends of key variable and product property predictions. Literature Cited 1. Friedman, Y.Z. Model-Based Control of Crude Product Qualities Hydrocarbon Processing, February 1994, pp Kesler, M.G. et al: New Method Dynamically Models Hydrocarbon Fractionation Hydrocarbon Processing, October 1995, pp Chapter 3 in the API Technical Data Book Petroleum Refining, March 1988 Page 7

10 FIG. 1: Preflash And Atmoshperic Columns Off-Gas TC Light Naphtha TC Heavy Naphtha Wild Naphtha Kero/Jet Hot Crude Furnace Oil Reduced Crude Wax Distillate FIG. 2: Crude Unit Operation Unsteady Crude Throughput Crude TBP slope Page 8

11 340 FIG. 3: Naphtha 90%-ASTM Temp (F) /3/00 1/8/00 1/13/00 1/18/00 1/23/00 1/28/00 2/2/00 FIG. 4: Kero 90%-ASTM Temp (F) /3/00 1/8/00 1/13/00 1/18/00 1/23/00 1/28/00 2/2/00 Page 9

12 FIG. 5: Furnace Oil 90%-ASTM Temp (F) /3/00 1/8/00 1/13/00 1/18/00 1/23/00 1/28/00 2/2/00 FIG. 6: Schematic View ONCE PER 10 MIN CPPM MEASUREMENTS (SUPERSET) ONCE PER MINUTE GCC MEASUREMENTS CONTROL Page 10

13 FIG. 6A: GCC Controls Off-Gas TC Light Naphtha TC Heavy Naphtha INFERENTIAL CONTROL Wild Naphtha Kero/Jet Preheated Hot Crude Furnace Oil Wax Distillate FIG. 7: System Overview Temp (F) Inferred TBP Vol% Solver/ Optimizer Calibrated T-T-T Model Column profile Tray # Temp (F) CPPM Data Conditioning Checking Results CPPM PI GCC INTERFACE DCS Page 11

14 FIG. 8: Naphtha Cutpoint Example Crude TBP slope Closed loop period Cutpoint Naphtha cutpoint Target FIG. 9: Inferred 90% vs Lab Data Furnace Oil Non Jet Target Lab Values Every 4 Hours Jet Operation Target (8/20/00 9/19/00) Page 12

15