Part II. Deployment of Continuous Online Monitoring of Local Tube Metal Temperature of a Refinery Depropanizer Reboiler

Paper from the AFRC 2016 conference titled Part II. Deployment of Continuous Online Monitoring of Local Tube Metal Temperature of a Refinery Depropanizer Reboiler

Continuous monitoring of local tube metal temperatures of process tubes is a difficult but desired objective for the safe and efficient operation of many process heaters and boilers.; A new measurement technology has been developed (see Part I. paper by the same authors presented at the 2015 AFRC meeting, Salt Lake City) for providing local tube metal; temperatures in such furnaces. With this paper we present the deployment and validation of this technology on the refinery process heater serving as a reboiler for the bottoms of a; depropanizer tower at the HollyFrontier Refinery in Woods Cross, Utah.; The technology brings together the existing process data acquisition on the process heater with high performance computer simulations of the combustion system coupled to; the thermodynamic properties of the multiphase process fluid, and with Bayesian learning from the historical data. By using these technologies together, an instrument model is; constructed to provide the local tube metal temperature throughout the radiant section of the reboiler. The continuous measurement of tube metal temperature is updated as the; existing measurements of operating variables (the process fluid temperature at the inlet and outlet of each coil in the heater, the process composition, process flow rates, and the; combustion stack gas temperature) are updated. The instrument model is calibrated and updated as new data are acquired.; This measurement technology has been tested offline on four months of historical data collected during the fall and winter of 2015. Subsequently, the method has been; hardened for deployment online, calibrated with optical data collected at the refinery, continuously validated with online process data, and deployed at the refinery. This paper; reports on the experience of testing, calibrating, validating, and deploying this technology.

CRSim A F R C 2 0 1 6 Part II. Deployment of Continuous Online Monitoring of Local Tube Metal Temperature of a Refinery Depropanizer Reboiler P. Smith: CRSim Inc. J. Thornock, S. Smith, M. Hradisky: The University of Utah D. Smith, P. Emett, M. Hardy: APCO Inc. K. Daines, B. Harris: HollyFrontier Corp. ABSTRACT Continuous monitoring of local tube metal temperatures of process tubes is a difficult but desired objective for the safe and efficient operation of many process heaters and boilers. A new measurement technology has been developed (see Part I. paper by the same authors presented at the 2015 AFRC meeting, Salt Lake City) for providing local tube metal temperatures in such furnaces. With this paper we present the deployment and validation of this technology on the refinery process heater serving as a reboiler for the bottoms of a depropanizer tower at the HollyFrontier Refinery in Woods Cross, Utah. The technology brings together the existing process data acquisition on the process heater with high performance computer simulations of the combustion system coupled to the thermodynamic properties of the multiphase process fluid, and with Bayesian learning from the historical data. By using these technologies together, an instrument model is constructed to provide the local tube metal temperature throughout the radiant section of the reboiler. The continuous measurement of tube metal temperature is updated as the existing measurements of operating variables (the process fluid temperature at the inlet and outlet of each coil in the heater, the process composition, process flow rates, and the combustion stack gas temperature) are updated. The instrument model is calibrated and updated as new data are acquired. This measurement technology has been tested offline on four months of historical data collected during the fall and winter of 2015. Subsequently, the method has been hardened for deployment online, calibrated with optical data collected at the refinery, continuously validated with online process data, and deployed at the refinery. This paper reports on the experience of testing, calibrating, validating, and deploying this technology. 1 of 11 INTRODUCTION Most modern instruments use indirect measurement. For example when a thermocouple is used in a suction pyrometer to measure the temperature of combustion gases it requires the construction of a somewhat elaborate model of the the environment in which the measurement is being made in order to convert the measured voltage to the temperature output. This model involves an understanding of the Seebeck, Thomson and Peltier effects to relate the temperature to the voltage or electromotive force. It also requires models of the radiative losses from the thermocouple bead to the various layers of the suction pyrometer. Conductive/convective heat transfer models are incorporated to relate the bead temperature to the gas temperature. In the end, the voltage that is the direct measure is converted through the model to the indirect measure of temperature. In this paper we refer to this type of indirect model as the instrument model. Instrument models require validation and calibration. The validation process quantifies the degree of uncertainty of the output of the instrument model (i.e. gas temperature) to the uncertainty in the instrument model input (i.e. the measured voltage). The calibration process reduces the uncertainty in model parameters based on independently acquired information (including uncertainty) about the instrument model output (i.e. gas temperature). The instrument model is then run in near real time, coupled with modern data acquisition systems to continuously produce indirect output measurements. In this example the instrument model reads voltages from the thermocouple and produces continuous near real-time indirect measurement of the gas temperature. High performance computing and Bayesian methods combine to provide an opportunity to efficiently use much more sophisticated instrument models than have ever been used before to produce indirect measurements of ever more difficult quantities of interest. In a companion paper1 by the same authors, we described a project to produce a continuous, indirect measurement of fireside skin temperature of process tubes in process heaters and boilers. Continuously monitoring local tube skin temperatures for all tubes across all tube lengths is a difficult but desired objective for the safe and efficient operation of many process heaters and boilers. In this paper we present the deployment of this technology on a refinery process heater serving as a reboiler for the bottoms of a depropanizer tower at the HollyFrontier Refinery in Woods Cross, Utah. 1 Continuous Online Monitoring of Fireside Tube Skin Temperature in a Depropanizer Reboiler, American Flame Research Committee (AFRC) annual meeting, Salt Lake City, UT, September 2015. 2 of 11 The development and deployment of this instrument followed a four phase work plan: • Phase 1) Develop the instrument model for the reboiler; thus, producing a proof-ofconcept for the tube skin temperature measurement. • Phase 2) Compare static skin and process fluid temperature measurements for HollyFrontier refinery; thus, calibrating and validating the instrument model. • Phase 3) Develop map of operating conditions, using historical HollyFrontier refinery data running offline; thus, testing the dynamic tube skin temperature measurements. • Phase 4) Deploy the online dynamic measurement at the refinery as part of the control system. The paper previously referenced described phase 1 (development and proof of concept). In this paper, after a descriptive review of the specific reboiler being instrumented, we discuss phases 2-4: validation, calibration, testing and deployment. THE REBOILER This fuel gas fired furnace acts as a reboiler on the bottom of the depropanizer tower.  The furnace adds 25.6 MMBTU/hr to the bottom of this tower.  It was designed for about 65 wt. % evaporation (91,750 lb/hr liquid, 176,000 lb/hr vapor) of charge passing through it at design conditions with an efficiency of 85%. The design maximum pressure is 500 psig at 450oF temperature.  The reboiler furnace normally operates at 285 psig at 355oF at the inlet and 270 psig at 424-440oF at the outlet. This is a cylindrical, vertical type furnace and is fired from four John Zink PSFG-16RM staged fuel gas burners with individual gas pilots.  The burners are mounted in the floor of the furnace.  There are 120 tubes in the depropanizer reboiler heater.  There are 64 finned and 16 bare or shield, horizontal tubes in the convection section.  The shield tubes are located below the finned tubes to give a more even distribution of heat to the product and shield the finned tubes in the convection section.  The 40 radiant tubes are 24'-0" long except the radiant inlet tubes which are 27'-6" long and the radiant outlet tubes which are 26'-6" long.  The radiant tubes are arranged vertically around the wall of the furnace. The reboiler heater has four identical coil passes.  Each pass is equipped with transmitters to show the individual coil flows, and exit process fluid temperature.   Each coil pass has 16 finned convection tubes, 4 shield tubes, and 10 tubes in the radiant section.  All the tubes are 4.5" O.D. A-106 Gr. B stainless steel material.  All the return bends are welded.  The 3 of 11 flow enters each coil pass on the top row of the convection section and flows down through the convection sections, out the cross over and into the top of each radiant section.  Flow enters each radiant coil pass, flows through the 10 tubes, and out the top of the radiant section and returns to the depropanizer tower.  The radiant section north west coil pass and north east coil pass outlets are on the north side of the furnace.  The radiant section south west coil pass and south east coil pass outlets are on the south side of the furnace. The furnace is 10'-5" diameter and 35'-0" tall, excluding the stack.  The stack is 4'-0" diameter and 82'-0" tall and rests upon a transition section and rectangular convection section on top of the cylindrical part of the furnace.  A damper is provided in the furnace stack and is manually operable from ground level.  Smothering steam lines are provided for the firebox.  The combustion air to the furnace is controlled by a damper system in the wind box on the bottom of the furnace.  A flue gas sample connection is provided in the stack below the damper for running flue gas analysis (for pollution control and efficiency). Figure 1: Photograph and schematic cross section of the HollyFrontier refinery AlkyDepropanizer Reboiler showing the 4 coil locations. 4 of 11 T H E VA L I DAT I O N The instrument model development for the continuous indirect measurements of the coil skin temperatures has been described previously. Here we describe the validation process and results. This validation process occurs concurrently with the continuous measurement and draws from data already being collected at the HollyFrontier Woods Cross refinery through the existing Honeywell distributed control and data acquisition system (dcs). Specifically, 25 continuous data points were used as the ‘primary' data for continuous updating of the SimMapp™ measurements. Those data are: • Fuel feed rate [Mscf/day] • Radiant section temperature [F] • Convective section temperature [F] • Stack gas temperature [F] • Process fluid vol. flow rate [BPH] • Process fluid flow rate for each of the four coils [BPH] • Alkylate and Butane to storage [BPH] • Mix Drum Analyzer hydrogen sulfide, hydrogen, air, methane, carbon dioxide, ethylene, ethane, propane, propylene, C4+ measurement [mol%] • Mix Drum Analyzer un-normalized total [mol%] • Mix Drum Analyzer gross heating value [BTU/scf] • Mix Drum Analyzer net heating value [BTU/scf] • Fuel Gas Mix Drum Temp [F] Two major sources of uncertainty contribute to measurement error: instrument model parameter uncertainty and measurement uncertainty in the 25 measured inputs to the instrument model. Both of these sources of uncertainty are propagated through the instrument model to produce a nonlinear response in the output of the instrument model. This uncertainty in the tube metal skin temperature is then reported as part of the measurement of the quantity of interest. Six outputs from the instrument model are measured continuously at the refinery and available through the dcs: • Process fluid temperature IN [F] • Process fluid temperature OUT for each of the four coils [F] • Mixed Process fluid temperature OUT [F] These measurements offer the opportunity to use Bayesian methods to learn from the data to reduce the uncertainty in the tube metal skin temperature by narrowing the uncertainty 5 of 11 in the model parameters and in the 25 measured inputs. This learning algorithm allows the uncertainty to evolve as a function of the history of the reboiler. This also constitutes a validation process for the instrument model. A typical validated output for one coil at one point in time is shown below in Figure 2. Figure 2: Tube metal skin temperature (solid red line) as a function of tube length from inlet to outlet of the southwest coil in the reboiler. The process fluid temperature (solid green line) and the vapor fraction (quality) of the process fluid (solid blue line) are also shown. The dotted lines around each solid line shows the uncertainty in the measurements at this point in the learning algorithm. The red circle shows the measurement of the process fluid outlet temperature as reported by the refinery dcs. The maximum skin temperature, its uncertainty and the timestamp for this coil are all listed in the figure title. The behavior of the skin temperature as a function of its length is discussed in the next section. 6 of 11 In T H E C A L I B R AT I O N Report process Date described above 8/7/2015 While the validation ensures that the instrument model is consistent with the dcs data collected by the refinery for the reboiler, it does not uniquely define the tube metal skin temperatures. The heat transfer coefficient for each coil requires Company HollyFrontier Customer an independent calibration. This calibration was performed by drawing on FLIR data collected at the refinery for each coil. The average internal transfer coefficient is Address 1070coil west 500 heat south, Site constrained to be consistent with the FLIR data. Figure 3 showsUtah an example of the FLIR West Bountiful, Address data taken on and thus corresponding to the 84087 calibrated coil tube metal skin temperature shown in Figure 2. Thermographer Joe Perlac Image and Object Parameters Contact Person Text Comments Figure 3: FLIR image taken of the bottom section of the southwest coil on 8/7/2015 @ 7:55. These data correspond to the SimMapp™ measurements of the same date and time Camera Model P50 F NTSC shown in Figure 2. Image Datefor the bottom8/7/2015 7:46:43 AM and top section The FLIR data include images (shown here), midsection (not shown) of each of the four coils. The SimMapp™ measurement and the FLIR data Image Name IR_0288.jpg both show quantitative and qualitative agreement regarding the location coolest temperatures forEmissivity the skin being at the top of 0.96 each coil at the firing rate for which these data were collected. As the process fluid moves Reflected apparent 68.0from °F the top to the bottom, the skin temperature rises, reaches a peak somewhere around 1/3 the distance from the bottom of temperature Object Distance Description 6.6 ft 7 of 11 8/7/2015; 7H3, Charge rate @ 1099. Inlet temp at 423deg. Outlet temp. at 467deg the furnace. Then the skin temperature drops. The skin temperature rises through the bend at the bottom of the furnace reaching a local maximum in the middle of the bend where the view factor for seeing the hot bottom of the furnace peaks. As the process fluid rises through the next section of the coil and moves toward the top, the skin temperature rises too. Again, it reaches a peak at a distance somewhere around 1/3 the distance to the top of the furnace, than the skin temperature drops and reaches a minimum as it moves into the bend at the top of the furnace. As at the bottom there is a peak in skin temperature midway through the bend. This process repeats itself through the ten vertical sections of each coil. At this calibration point the overall maximum skin temperature (711F) occurred partway up the fourth vertical section of this southwest coil. The location of the hotspot varies from location to location and from coil to coil as the operation of the reboiler changes. The magnitude of these variations are discussed next as learned from the offline testing phase of this project. 8 of 11 THE OFFLINE TESTING This SimMapp™ instrument was installed and interfaced to a SQL database which was continuously populated with the refinery dcs data in increments of 5 minutes covering the period of 10/1/2015 to 12/31/2015. This three month online test allowed for software shakedown, error handling for a range of operating conditions, data conditioning for individual dcs data failures, etc. Over this three month period of historical data testing, the maximum skin temperature as a function of time for the southwest coil is shown in Figure 4. Figure 4: History of three month offline testing showing the SimMapp™ maximum skin temperature for the southwest coil (magenta) and the corresponding process fluid temperature (green). During this offline testing the SimMapp™ instrument produced a figure like that shown in Figure 2 every 5 minutes (the data rate for the input data set from the dcs) for each of the 4 coils in the reboiler. The hardware and software of the instrument were configured to meet specification required by HollyFrontier; thus, proving readiness for online deployment, which is discussed in the next section. 9 of 11 THE DEPLOYMENT In August of 2016, the instrument was deployed online at the HollyFrontier Woods Cross refinery. The data flow for the instrument in online mode is shown in Figure 5. Holly's DCS polls data from the CRSim modbus server for computed data Holly's DCS Broker writes computed data from SQL into the CRSim modbus server Modbus Server Holly's DCS writes to the CRSim modbus server raw values needed for computation Broker Software Broker reads raw data from CRSim modbus server Broker software reads computed data from SQL Broker writes raw data into SQL 'alarms' Table 'modbusmap' Table 'brokersettings' Table 'sim' Table 'calibration' Table SQL Tables 'SimMapp' DB 'simsettings' Table 'dcs' Table SimMapp reads raw data from SQL and computes it SimMapp writes computed data into SQL SimMapp Analysis Software CRSim Webserver SimMapp publishes a '.png' file of the latest 4 coil profiles to the CRSim Webserver Figure 5: Data flow schematic for the SimMapp™ instrument online at the HollyFrontier Woods Cross refinery. 10 of 11 CONCLUSIONS We have used high performance computing (hpc) and Bayesian analysis methods together to provide a new approach to instrument modeling for indirect measurements of fireside skin temperature of process tubes in process heaters and boilers. With this new measurement tool (called SimMapp™), we continuously and online track local tube skin temperatures for all tubes across all tube lengths. The value of this instrument is found in improved safety and efficiency in the operation of process heaters and boilers. Calibration, validation, offline testing, and online deployment of this technology has been completed at the HollyFrontier Woods Cross refinery alky-depropanizer reboiler. These measurements, along with their uncertainties, provide information heretofore unavailable to the refinery. The continuous monitoring of the location and magnitude of the maximum tube metal temperature is being considered for integration into operational procedures for the refinery. 11 of 11