COKER DASHBOARD USER MANUAL - What Your Coke Drum Is Telling You

1. Introduction

1.1 What Is SPM for Cokers

Structural Performance Management (SPM) is the system Akselos provides. It continuously calculates the real structural limits of critical infrastructure from operating data, so you can safely increase throughput, reduce downtime, and extend asset life. It does this with physics-based AI and high-fidelity structural twins that show how the steel actually responds under real conditions.

At its core is Reduced-Basis Finite Element Analysis (RB-FEA), Akselos’ patented solver. It runs high-fidelity structural analysis 1000× faster than conventional FEA and just as accurately, so a structure can be assessed continuously rather than in one-off studies.

SPM connects operating data, inspection records, and design information into one live structural view. For a coke drum that means engineers and operators can see how the drum responds as conditions change, test operating and maintenance strategies before committing to them, and move decisions from conservative assumptions to evidence.

This article is the user manual for the Coker Dashboard, the report where SPM for Cokers presents its results for one coke drum. It covers what each page reports, how to read it, and what each number mean. How SPM for Cokers turns your plant data into those results is outlined next.

1.2 How SPM for Cokers Works

Three parts sit between your plant data and the numbers you read on the dashboard: the stages that carry the data, the structural model of your drum, and the three assessments run on it.

1.2.1 From Plant Data to the Coker Dashboard

Akselos SPM for Cokers is SPM applied to the coke drums of a delayed coking unit. It runs in three stages between your plant data and the Coker Dashboard, and Figure 1 shows how they connect.

• The structural model. Akselos engineers build a high-fidelity finite element model of your drum from its design geometry, its materials, its boundary conditions and its load scenarios, with a sensor data connection supplying the operating history it is solved against.
• The assessments, on the Akselos Cloud Solver. Fatigue, bulging and crack are assessed on the structural model for every processed cycle, solved on multi-core cloud servers. RB-FEA accelerates large-scale models, finite element analysis and hybrid solvers are supported, and every data transmission is encrypted.
• The Coker Dashboard, on the Akselos Portal. Results reach you on the Coker Dashboard, a report in the Akselos Portal that opens in your browser, with nothing to install. The Portal is where simulation data is managed, teams work together and assets are monitored.

[IMAGE: image1.png]

Figure 1. The three stages of SPM for Cokers, from the structural model through the Akselos Cloud Solver to the Akselos Portal

1.2.2 The Structural Model of Your Drum

The structural model returns the stress at every assessed location of the drum, and it is solved on two shapes:

The model is divided into zones agreed with your own specialists, and the zone names are the ones printed in every ranked table.

The thermal load is set by the coking cycle. Its severe part is the quench: water is injected into a drum still at coking temperature and channels unevenly through the coke bed, so one side of the shell cools faster than the other. Figure 2 shows the drum at two moments of one cycle.

[IMAGE: image2.png]

Figure 2. One coking cycle, with the drum at the coking plateau and during the quench

1.2.3 Fatigue, Bulging and Crack Assessments

Three assessments run on the structural model, each answering its own question. Figure 3 shows where operating data and inspection data enter them.

[IMAGE: image3.png]

Figure 3. The assessment chain, from model preparation to reported life

• Fatigue: how much of the design fatigue life each assessed location has used, summed cycle by cycle from the recorded operating data on the design shape. Read it on Fatigue Status, Cycle Inspection and Historical Trends.
• Bulging: how far the shell has permanently deformed, and how close that has taken the material to its limit, reported as the plastic strain limit fraction, written PSLF. It runs on the scanned shape from the latest laser scan, and the same scan gives ovality, checked separately. Read it on Bulging Inspection.
• Crack: whether each tracked flaw is acceptable on a Failure Assessment Diagram, written FAD, and how long it stays acceptable as it grows. Flaws come from the inspection record, or are placed for screening where none has been recorded. Read it on Crack Status and Crack Inspection.

All three follow API 579-1/ASME FFS-1, 2021 edition, and bulging also draws on ASME BPVC Section VIII Division 2, 2019 edition.

Info: The methodology article covers how the recorded data is turned into loads, how the model is built, and how each assessment is computed, with the clause of each standard every step draws on.

2. Getting Started

2.1 Who This Guide Is For

Two roles use the dashboard, each from a different set of pages:

• Operators monitor the recorded sensor data and the exceedances of the design operating window on Process Monitoring.
• Structural Integrity engineers assess fatigue, bulging and crack and their trends on Fatigue Status, Crack Status, Bulging Inspection, Crack Inspection, Cycle Inspection and Historical Trends, as an input to turnaround planning and inspection scope.

2.2 Which Page Answers Your Question

Find your question in the left column, then open the page named beside it. Each page has its own section under Detailed Coker Dashboard Pages.

2.3 Sign In and Open the Dashboard

The Coker Dashboard runs in your browser as a report inside your organization on the Akselos Portal. You need two things before it opens: an Akselos Portal account, and access to your organization. Both start at the Akselos Portal sign-in page, shown in Figure 4.

[IMAGE: image4.png]

Figure 4. The Akselos Portal sign-in page

• An Akselos Portal account. If you do not have one, see the article on signing up for an Akselos account.
• Access to your organization, granted by an administrator. An organization is your company’s own space on the Portal, holding its assets and reports.

The dashboard is published as a report inside your organization, so you reach it through Reports. Figure 5 shows where it sits.

[IMAGE: image5.png]

Figure 5. A dashboard listed among the reports in an organization

The entry is named after your unit or asset, and an organization with several assets carries several reports.

1. Log in to the Akselos Portal.
2. Select your organization.
3. Open Reports.
4. Select the dashboard for your asset. It opens on Home, shown in Figure 6.

[IMAGE: image6.png]

Figure 6. The dashboard as it opens

Home is the landing page. Every other page is reached from the navigation panel on the left.

2.4 Move Between Pages

Every page shares one frame. A navigation panel runs down the left, and the page fills the area to its right. Figure 7 shows the panel with every module expanded.

[IMAGE: image7.png]Figure 7. The navigation panel with all modules expanded

The panel groups the eight pages under four modules. Expand a module to reveal its pages, and the page you are viewing stays highlighted while you read it.

Two controls sit at the foot of the panel. Other Assets moves you to another asset’s dashboard, where your organization runs more than one. Logout ends the session and returns you to the Portal.

Pages also open one another directly, through the card arrows on Home, Inspect Last Cycle on Process Monitoring, a double-clicked bar or dot, and the Cycle ID links. Figure 8 maps those routes, most of which end at Cycle Inspection.

[IMAGE: image8.png]

Figure 8. The shortcuts between pages, most of them ending at Cycle Inspection

2.5 Check That Plant Data Is Arriving

Data Connection

The dashboard header carries two status indicators on every page, shown in Figure 9. The first reports whether operating data is currently reaching the dashboard.

[IMAGE: image9.png]

Figure 9. The status indicators in the dashboard header, for the data connection and the drum state

Note: Every timestamp on this dashboard is shown in UTC. A dashboard time compared against a plant record kept in local time has to be converted first.

Drum State

A second indicator reports the working state of the drum itself. While operating data is not reaching the dashboard the state cannot be determined and reads Unknown. Once data arrives it reads In Operation or Not in Operation.

How Data Reaches the Dashboard

Operating data reaches the dashboard by one of two routes:

• A live connection. Your sensors feed your own data system, and that system is linked to the Akselos database. From there the assessment runs on its own, and the dashboard updates as new cycles complete. Building the link is joint work between your IT team and ours.
• A manual run. Akselos engineers take the operating data, prepare it, and load it for assessment. The results appear on the same dashboard pages.

Info: Both routes run the same assessments and produce the same pages.

To set up either route, or to move from one to the other, contact Akselos support at [email protected].

3. Features Every Page Shares

Every page shares the same controls. They are covered once here, and each page section refers back to them.

3.1 Widget Help, Selectors and Status Bands

Widget Help

Almost every widget carries a ? that opens a short description of what it shows. Figure 10 shows one open.

[IMAGE: image10.png]

Figure 10. The description a widget opens from its ? control

Cycle and Inspection Selectors

Some pages report against a selection you make, set from a selector at the top of the page. A cycle selector lists the recorded cycles, most recent first, and reports the start and end timestamps of the one you pick. An inspection dataset selector lists the inspection campaigns loaded for the drum. Figure 11 shows a cycle selector.

[IMAGE: image11.png]Figure 11. A cycle selector, with the start and end timestamps of the selected cycle

Note: Changing the selection re-reports the whole page. Before comparing a figure on one page against a figure on another, check that both report against the same cycle or the same inspection.

Status Bands

Several pages open with a status word rather than a number, and the same two bounds set it everywhere:

• NOMINAL, above the upper bound
• WARNING, between the two
• ATTENTION REQUIRED, at or below the lower bound

What changes from page to page is the metric measured against those bounds, and each page names its own.

3.2 Measured Values, Computed Values and Assessed Locations

Some numbers on this dashboard are measured. The rest are computed.

• Measured: the process variables your sensors record, read raw on Process Monitoring. Which variables appear depends on the sensors installed on your asset.
• Computed: stress, displacement, damage, the life estimates, and every fitness-for-service result.

The structural model computes at assessed locations. An assessed location is a single position on the drum wall, fixed by its azimuth around the circumference and its elevation up the vessel. Damage therefore arrives as a ranked list of locations, never as one number for the whole drum.

3.3 Charts: Hover, Zoom, Filter and Download

Hover Readout

Hovering on the plot lists the timestamp and the value of every visible series at that moment.

Sensors Selection

• The series list, at the top right, opens Sensors Selection. All shows every series, None clears them, and the checkboxes turn series on and off one at a time. Colors in the list match the traces. Figure 12 shows the chart filtered to one sensor.

[IMAGE: image12.png]

Figure 12. Sensors Selection, with the chart filtered to a single sensor

Idea: To compare a small number of sensors, press None first, then tick only the ones you want.

Range Sliders and Zoom

• Two range sliders narrow what is plotted: the one under the horizontal axis sets the dates, the one beside the vertical axis sets the values, and scrolling on the plot zooms.

Downloading the Data

The download control, beside the series list, saves the data behind the chart as a CSV file: a Timestamp column, then one column for each sensor on that chart, at the interval those sensors record on, for the period selected. Figure 13 shows the layout.

It is offered on the sensor charts of Process Monitoring and Cycle Inspection only.

[IMAGE: image13.png]

Figure 13. A downloaded chart file, with one column per sensor

3.4 Model Viewers and Color Scales

A result viewer draws a computed field onto the drum geometry, so you can see where a quantity concentrates instead of reading it out of a table. Figure 14 marks its controls.

[IMAGE: image14.png]

Figure 14. A result viewer, with the color scale, unit control and orientation indicator

Moving the Model

Scroll to zoom, hold Ctrl with the middle mouse button to pan, hold the middle mouse button alone to rotate, and use reset to return the model to where it opened.

Color Scale and Units

The color scale gives the value range, and the unit control sets the units it is reported in.

Note: A color scale rescales to whatever the page is reporting, so the same color means a different value under a different selection. Read the scale rather than comparing two views by color.

Orientation Indicator

An orientation indicator in the corner marks Vertical Up and the compass directions, which is what lets you turn a feature on screen back into the position coordinates the ranked tables report.

3.5 Flat Shell Maps

Three widgets draw the drum shell as a flat sheet: Crack Specification on Crack Status and on Crack Inspection, and Inspection Visualization on Bulging Inspection. A drum is a cylinder, and a cylinder can be cut down one line and rolled out flat without distorting it. Figure 15 shows the cut.

[IMAGE: image15.png]

Figure 15. The shell cut at North and rolled out flat, with one point on the wall and the same point on the sheet

Cut at North and unroll, and every point on the wall lands at one point on the sheet. Azimuth runs left to right from North through East, South and West and back to North, and elevation runs bottom to top on the same datum the ranked tables use. Horizontal bands mark the shell courses and the plate seams inside them, so a feature can be read against the welds nearest it.

4. Detailed Coker Dashboard Pages

This section covers the dashboard page by page, in the order of the navigation panel. Each widget is described by what it shows and how to read it, under the name the dashboard prints.

4.1 Home (Landing Page)

Home gives the overall condition of the drum and the assessment setting the limit, with the current condition, the design data and the model below it. Figure 16 shows the page.

[IMAGE: image16.png]

Figure 16. The Home page

4.1.1 Overall Coker Analytical Horizon

Overall Coker Analytical Horizon reports, in years, the shortest result of the three assessments, fatigue, bulging and crack, separately for the Pressurized and the Non-pressurized components. Figure 17 shows the panel.

[IMAGE: image17.png]

Figure 17. The Overall Coker Analytical Horizon panel, reported for both component groups

Each group carries one bar per assessment, crack, bulging and fatigue, with the years that assessment allows. A MIN marker sits on the bar setting the headline figure, and the component named beneath that figure is where the governing result was found. An assessment with no loaded result reads No data rather than zero, so a missing dataset never produces a short headline figure.

The status word above the figure follows the bands in Widget Help, Selectors and Status Bands:

• Above 25 years reads NOMINAL
• Above 5 years and up to 25 years reads WARNING
• 5 years or less reads ATTENTION REQUIRED

Note: The headline years are the shortest of three assessments, not an average and not a fatigue figure on their own. Read the MIN marker to see which assessment produced it, then open that page.

4.1.2 Current Condition Cards

Six cards report the current condition, each with a colored edge carrying its status. Figure 18 shows the row.

[IMAGE: image18.png]

Figure 18. The six current condition cards, each with a colored edge carrying its condition

• Temperature and Pressure report the last recorded reading, the sensor that supplied it, and the time it was taken. Each state whether that reading sits inside the design operating window, printed as Inside DOW, and names the design limit it was checked against.
• Utilized Cycles reports the cycles recorded as a percentage of the design cycle limit, with both counts stated beneath.
• Bulging reports the highest plastic strain limit fraction twice, for the Shell Body and for the Cone.
• Crack reports the length and the depth of the governing flaw, each with its component. A length of 0.100 m with a depth of 0.001 m is a screening flaw placed by the assessment rather than a measured one, as Crack Status explains.
• Fatigue reports the highest damage limit fraction once per component group, each figure with the component carrying it.

Each card carries an arrow at its corner that opens the page holding the detail behind that number.

4.1.3 Equipment Data

Equipment Data lists the asset the dashboard reports on and the data every assessment runs against. Most rows are design data, fixed when the deployment is built. Two rows move.

The two design conditions are the same limits used elsewhere on the dashboard. They set the Inside DOW check on the Temperature and Pressure cards, and the design reference line drawn on the Process Monitoring charts.

4.1.4 Model

Model shows the structural model every assessment runs on, divided into the zones your results are reported against. Figure 19 shows the panel.

[IMAGE: image19.png]

Figure 19. The Model panel, showing the assessment zones and the elevation extents

The panel holds two views of the drum: one colored by zone, using the zone names printed in every ranked table, and one plain, marked with the elevation extents of the model.

Elevations are measured from the bottom tangent line at 0 m, positive above it and negative below, and every elevation on the dashboard uses that datum. The axis indicator marks vertical, north and east, the directions azimuth is measured from.

4.2 Process Monitoring

Process Monitoring shows the recorded plant data every fatigue result is computed from, against the design operating window. Figure 20 shows the page.

[IMAGE: image20.png]

Figure 20. The Process Monitoring page

Inspect Last Cycle, in the page header, is a shortcut to Cycle Inspection, where one cycle is broken down on its own.

4.2.1 Time Information

Time Information governs the whole page. Set the span with the Duration dropdown, or set it directly with the From date and To date fields and the slider beneath them. Every card and every chart re-reports against whatever you set. Latest Update beside the panel gives the timestamp of the most recent data the dashboard holds.

The Duration list offers Last 7 days, Last 1 month, Last 3 months and Last 6 months. Custom hands the span back to the From date and To date fields. Figure 21 shows the list open.

[IMAGE: image21.png]

Figure 21. Time Information, with the Duration list open

4.2.2 Process Charts

Each measured variable gets its own time-series chart, plotting every sensor of that type as a separate trace over the period you selected. Figure 22 shows the four charts.

[IMAGE: image22.png]

Figure 22. The four process charts, with the Max Design line on Thermocouples and Pressure Sensor

Note: The units on these charts are fixed. There is no unit selector, so a value read here is always in DegC, %, Barg or m3/h.

Note: A coke level reading can pass 100%. The percentage is reported against the elevation at which the level instrument sits, not against the height of the drum, so coke standing above that elevation reads above 100 without the drum being full.

Thermocouples and Pressure Sensor each carry a Max Design line, set from the design condition in Equipment Data and checked by the cards below. Coke Level and Flow Rate Sensor have no design condition, so they carry no line.

Charts work as described in Charts: Hover, Zoom, Filter and Download, and a download covers the period you selected, one file per chart. Figure 23 shows the layout of one file.

[IMAGE: image23.png]

Figure 23. Downloaded data from a Process Monitoring chart

Info: Which charts appear follows the instrumentation on your asset. A sensor type your plant does not meter carries no chart here.

4.2.3 Maximum Value Cards

Four cards, Max. Temperature, Max. Pressure, Max. Coke Level and Max. Flow Rate, report the highest value in the period you selected and the sensor that recorded it.

The temperature and pressure cards also read Inside DOW or Exceeds DOW against the stated limit, and carry two counters, 30d exc and 90d exc, giving the recorded timesteps above the limit in the last 30 and 90 days.

Note: The maximum value follows the period you selected. The two exceedance counters do not. They always look back 30 and 90 days from the most recent data, whatever the period is set to, so a card can show a short-period maximum next to a 90-day exceedance count.

4.2.4 DOW Exceedance Log

DOW Exceedance Log lists the individual excursions behind the counters on the cards. Switch between the Temperature and Pressure tabs to choose the parameter, and the subtitle restates which parameter you are reading, the DOW limit applied, and the window covered. Figure 24 shows the log.

[IMAGE: image24.png]

Figure 24. The DOW Exceedance Log, with the parameter tabs and the exceedance column

Each row gives the Date/Time of the excursion, the Sensor Name that recorded it, the Recorded Value and the Exceedance, which is the amount by which that reading passed the limit. Where nothing has gone outside the window the log says so rather than showing an empty table.

Note: The log covers a fixed 90 days and ignores the period set in Time Information. Selecting a longer span does not lengthen the log, and selecting a shorter one does not shorten it.

4.2.5 Sensor’s Location

Sensor’s Location places every instrument tag on the vessel model, so a tag named on a card, in a chart legend or in the exceedance log can be traced to where it physically sits on the drum. Zoom, pan, rotate and reset work as described in Model Viewers and Color Scales.

Tag names follow your site convention. In the convention used here a temperature indicator carries TI, a pressure indicator PI and a level indicator LI, and where a tag does not follow that pattern, your own tag register is the reference.

Use the panel to confirm where a sensor sits and which sensors share its elevation. Between instrumented positions the temperature applied to the model is interpolated rather than measured.

The next two pages sit under SPM Monitoring in the navigation panel. Where Process Monitoring shows what the sensors recorded, these report what the structural model computed from it.

4.3 Fatigue Status

Fatigue Status answers how much of the drum’s design fatigue life has been used, how fast it is being used, and which locations are using it. You reach it by expanding SPM Monitoring in the navigation panel, where it sits alongside Crack Status. Figure 25 shows the page.

[IMAGE: image25.png]

Figure 25. The Fatigue Status page

Every widget on this page reports twice, once per component group. The two groups carry different loads and are never combined:

• Pressurized covers the pressure-retaining components, such as the shell, cone and head.
• Non-pressurized covers the components not under internal pressure, principally the skirt. The junction where the skirt meets the shell is pressure-retaining, so it is grouped with the pressurized components rather than with the skirt.

Note: Damage only accumulates. Gentler operation slows the rate at which a location gains damage, and the total already reached stays where it is.

4.3.1 Status Cards and Cycle Selection

Two cards head the page, one per group, each opening with a status word set by the bands described in Widget Help, Selectors and Status Bands. The metric behind the word here is Min. Est. Fatigue Life. Below it each card reports:

• Max. Damage, the highest fatigue damage limit fraction reached in that group
• Last Cycle Damage, the fraction contributed by the most recent cycle
• Min. Est. Fatigue Life, the shortest estimated fatigue life in that group, extrapolated from how the last cycle ran

Each card also names the location governing those figures and the cycle they belong to.

Min. Est. Fatigue Life extrapolates from the most recent cycle, so one unusually severe cycle shortens it sharply and one mild cycle lengthens it. Read it as the rate the last cycle implies, and read Max. Damage beside it for the accumulated condition.

Info: Min. Est. Fatigue Life covers fatigue only. The Overall Coker Analytical Horizon on Home is the wider metric taking crack and bulging in as well, which is why the two can differ.

Cycle ID at the top right selects which cycle the page reports. The list holds every cycle that has been analyzed, most recent first, and the field beneath reports that cycle’s start and end timestamps, marked (Latest) for the newest. Changing it re-reports every widget on the page.

4.3.2 Est. Fatigue Life at Pressurized and Non-pressurized Parts

Est. Fatigue Life at Pressurized Parts and Est. Fatigue Life at Non-pressurized Parts put the figure from the cards onto a gauge, so the number reads against its bands rather than on its own. The scale is fixed in the product: red up to 5 years, yellow from 5 to 25, green above 25. Where the estimate runs past the top of the scale the needle rests at the end and the value reads as greater than the maximum.

4.3.3 Top 10 Damage Locations

Two ranked tables list the ten highest-damage locations in each group, with the position columns used on every later page. Figure 26 shows the component key that opens from either table.

[IMAGE: image26.png]

Figure 26. The component key, opened from the table tooltip

The tooltip on either table opens this key: each zone named against its elevation band, with the azimuth and elevation conventions.

Note: The two tables are ranked independently and the groups sit at different magnitudes, so rank 1 in one table is not comparable to rank 1 in the other. Read them together, because the location governing the whole asset may sit in either.

A rank orders locations by damage. It does not measure the interval between one rank and the next.

The locations at the top of these tables are the ones structural integrity engineers weigh first when the scope of a turnaround is being set.

4.3.4 High Fatigue Damage Locations

High Fatigue Damage Locations puts the two tables onto the drum itself, so a row becomes a point you can see rather than a pair of coordinates you have to picture. Figure 27 shows both sets of markers.

[IMAGE: image27.png]

Figure 27. High Fatigue Damage Locations, with both sets of ranked markers

Twenty markers are plotted, the ten from each table, and the prefix and color say which table a marker came from. Markers prefixed P# are pressurized locations, drawn in red. Markers prefixed NP# are non-pressurized, drawn in yellow. The number after the prefix is the rank in that table, so P#1 is the first row of the pressurized table.

Zoom, pan, rotate and reset work as described in Model Viewers and Color Scales.

4.3.5 Per-Cycle Damage Statistics

Per-Cycle Damage Statistics reports what each cycle cost rather than the running total, so an abnormal cycle stands clear of the cycles around it. Figure 28 shows the chart and its controls.

[IMAGE: image28.png]

Figure 28. Per-Cycle Damage Statistics, with the granularity, processing and date range controls

Granularity sets what one bar covers:

• Cycles, one bar per cycle, for finding a single abnormal cycle in the record
• Months, one bar per month, for seeing whether a period of operation was harder than those around it
• Years, one bar per year, for comparing one year of operation against another

Processing sets how the cycles inside each bar are reduced to one number:

• Max, the worst single cycle in that period, for locating the peak
• Avg, the mean across that period, for judging how severe operation typically was
• Sum, the total accumulated in that period, for reading how much life the period consumed

With Granularity set to Cycles each bar already covers one cycle, so Processing has nothing left to combine.

Max and Avg can diverge when operation changes: the worst cycles can rise while the average, held down by the many normal cycles, barely moves.

Sum is comparable only where bars cover equal, complete periods. A part month, or a month holding a shutdown, reads low because it holds fewer cycles.

From and To set the window, and Apply commits it. The window does not follow the field as you type it.

An orange Average Damage line runs across the plot with its value labeled, recalculated for whatever combination you are viewing, so a bar standing above the line is above average for the view you are in.

Idea: To open a cycle that stands out, zoom in with the range sliders until you can read its Cycle ID, then double-click the bar. The dashboard opens that cycle in Cycle Inspection, where the process and structural detail behind it is broken down.

4.3.6 Fatigue Life Prediction

Fatigue Life Prediction carries the damage curve forward. The solid historical curve ends at the present, and a projection continues from that point for the scenario selected, each scenario assuming future operation repeats a chosen stretch of past operation. Figure 29 shows the chart.

[IMAGE: image29.png]Figure 29. Fatigue Life Prediction, with the historical curve, a projection and the reference lines

Five scenarios are available, built from the last cycle, the last 10, 30 and 90 cycles, and the 10 worst cycles. Each projection is labeled at the date it meets the failure line. Scenario Info opens a table giving, for every scenario, the estimated date, the annual rate and the estimated life with an approximate cycle count. The Pressurized and Non-pressurized toggle switches which group is projected.

Two reference lines are drawn. Failure 100% marks full consumption of design fatigue life, and that line is the code basis. Maintenance (80%) marks a margin below it, set with the asset owner rather than by a standard. The 80 percent this build carries is the value agreed for your asset, not a code criterion and not a figure this article recommends.

Note: A projection extrapolates a past operating pattern forward. It describes where the current pattern leads, not when the asset will fail, and it moves whenever operation changes.

4.3.7 Top 10 Best Cycles and Top 10 Worst Cycles

Two tables close the page: Top 10 Best Cycles, the cycles that consumed the least fatigue life in one cycle, and Top 10 Worst Cycles, those that consumed the most. Both carry Rank, Cycle ID, Start Date and Fatigue Damage Increment (%), and both rank across the full recorded history rather than any selected period.

Each Cycle ID is a link that opens that cycle in Cycle Inspection, the quickest route from a high-damage cycle to the operating data behind it.

4.4 Crack Status

Crack Status answers how long each tracked flaw has before it reaches its limit size, and which flaw governs that figure. It sits under SPM Monitoring alongside Fatigue Status, and like that page it reports against one cycle you select. Figure 30 shows the page.

[IMAGE: image30.png]

Figure 30. The Crack Status page

Info: Crack Status predicts, on every cycle, how long each tracked flaw has left. Crack Inspection, under Asset Integrity, reports the flaws one inspection campaign recorded.

The table holds two kinds of flaw: flaws an inspection recorded, and screening flaws the assessment places where nothing has been recorded, at the locations the bulging assessment identifies. The tooltips name the screening dimensions Bounding Crack Length and Assumed Crack Depth, and describe them as a conservative screening model assumption based on the API 579 code rather than a confirmed field measurement.

Note: A row showing a crack length of 0.100 m and a depth of 0.001 m together is a placed flaw, not a measured one, and nothing else in the table marks it as such. Read the estimated life on such a row as a screening figure for a crack that has not been found, not as the life of a flaw on the drum.

4.4.1 Status Cards and Cycle Selection

Three cards head the page.

Overall Status gives the condition of the drum’s crack population in one word, set by the bands described in Widget Help, Selectors and Status Bands. The metric behind the word here is the minimum estimated crack life, measured against the same bounds the other assessments use.

Minimum Estimated Crack Life reports the shortest estimated time until a flaw grows to its limit size, and names the flaw and zone it belongs to. That figure is the smallest value in the Estimated Crack Life column of the table below, lifted to the header so the governing flaw is visible before the table is read.

Cycle ID selects the cycle the page reports, and gives that cycle’s start and end timestamps beneath.

4.4.2 Crack Details

Crack Details lists every flaw the page is tracking, with where it is, how big it is, and whether it passes.

Lr and Kr, the two ratios that place each flaw on the FAD Curve, appear as columns.

PASS means the flaw still sits inside the acceptable region of the diagram, and FAIL is its counterpart. The tooltip opens the same component key used by the fatigue tables, so a row here reads against the same zones and the same elevation datum.

4.4.3 FAD Curve: API 579 Part 9 Level 2

The FAD Curve: API 579 Part 9 Level 2 widget plots each flaw on a Failure Assessment Diagram, which sets two ratios against each other. Lr compares the applied load with the load that would cause plastic collapse, and Kr compares the applied stress intensity with the fracture toughness of the material. Figure 31 sets out how the diagram is read.

[IMAGE: image31.png]

Figure 31. How a Failure Assessment Diagram is read, with one flaw growing from its current size toward the boundary

A point under the curve is acceptable against fracture and against collapse at the same time. Partial safety factors are applied before a point is plotted, so its position already carries a margin.

Lr runs along the horizontal axis and Kr up the vertical. The FAD Curve is the boundary, the shaded Acceptable Range sits beneath it, and a point inside the shaded region is acceptable while a point outside it is not. A vertical dashed Cut-off line marks the plastic collapse limit.

Three marker types are plotted for each flaw, and Figure 32 shows them on the widget:

• Current Crack Size, the flaw as it stands today
• Developing Crack Size, tracing where continued growth takes it
• Critical Crack Size, the size at which it reaches the boundary

[IMAGE: image32.png]

Figure 32. The FAD Curve, with the acceptable range and the three crack size markers

Hovering a marker reports its Lr, Kr, crack length and crack depth, and the critical marker adds the critical length and depth.

Info: The Cut-off value is computed for the flaw and the cycle being assessed rather than fixed once, so it differs between locations and between pages.

4.4.4 Crack Specification

Crack Specification takes the cylindrical shell and lays it flat, so every flaw can be seen at once rather than one face at a time. The sheet reads as described in Flat Shell Maps, and Figure 33 shows it.

[IMAGE: image33.png]

Figure 33. The Crack Specification shell map, with the shell courses and plate seams

Each flaw is plotted as a marker carrying its identifier, against the shell courses and the labeled plate seams, which is how a row in Crack Details becomes a position on the drum.

The remaining four pages sit under Asset Integrity in the navigation panel. Each one narrows the view: to a single cycle, or to a single inspection campaign.

4.5 Cycle Inspection

Cycle Inspection explains a single cycle. Where Fatigue Status reports the cumulative picture, this page shows what the plant did during one cycle and what the structural model computed from it. Figure 34 shows the page.

[IMAGE: image34.png]

Figure 34. The Cycle Inspection page

4.5.1 Cycle Selection

Cycle Selection sets the cycle the whole page reports. Enter a cycle number directly or step through with the arrows, and the panel gives that cycle’s start and end timestamps.

Beneath the number, Status reads Good or Bad. The classification compares the maximum single damage in the selected cycle against the averaged single damage of all recorded cycles: higher than average reads Bad, lower reads Good. The tooltip shows both figures used in the comparison. The average covers the full recorded history and is recomputed as each new cycle is added.

Note: Good and Bad compare one cycle against the fleet of cycles this drum has run. The comparison is statistical rather than a code criterion: a Good cycle means lower than this drum’s average, not that the cycle sat within any limit.

4.5.2 Cycle Peak Values

Four cards report the peaks reached during the selected cycle. The first pair is what the plant measured, and the second pair is what the structural model computed from those measurements.

Von Mises stress combines the full three-dimensional stress state at a point into one number, which is what allows a computed stress to be compared against a material limit.

Fatigue damage follows the stress range a cycle produces rather than its peak, so two cycles with the same Max. Von Mises Stress can cost very different amounts of life.

4.5.3 Sensor Data of Cycle

Sensor Data of Cycle plots the raw readings across the selected cycle, with the horizontal axis bounded by that cycle’s start and end. A dropdown in the header switches the variable, and the options follow the sensors your asset carries: temperature, liquid level and pressure at minimum, with any further instrumented variable joining the list.

The hover readout, the Sensors Selection list, the range sliders and the download control all work as described in Charts: Hover, Zoom, Filter and Download. Downloading here gives you the readings for that one cycle rather than a whole period. Figure 35 shows the layout.

[IMAGE: image35.png]

Figure 35. Downloaded data from Sensor Data of Cycle, bounded by the selected cycle

4.5.4 Displacement of Top Drum

Displacement of Top Drum is a polar plot showing how far the top of the drum moved sideways during the cycle, and in which direction. Figure 36 shows the plot.

[IMAGE: image36.png]

Figure 36. Displacement of Top Drum, with the displaced position plotted against the compass axes

North, south, east and west mark the reference axes, and the plotted point sits at the displaced position. The panel reports the lateral displacement as north and east components in meters, the vertical displacement in meters, and the location the figures belong to with its elevation. Hovering the point repeats the components.

This is one of the measurements that tracks the Banana Effect: uneven quench cooling around the circumference contracts one side of the shell faster than the other, which bows the upper vessel to one side and concentrates fatigue damage at the skirt-to-shell junction. It is a different effect from ovality, the departure of a cross-section from round, which is read from the laser scan on Bulging Inspection.

4.5.5 Total Single Fatigue Damage by Elevations

Total Single Fatigue Damage by Elevations is a bar chart with elevation along the horizontal axis and the single-cycle damage up the vertical. Each bar carries its value, and an elevation that took no damage in the cycle reads zero. The tallest bar is the elevation that governed the damage for that cycle, which is the fastest way to tie a cycle back to a place on the drum. Its tooltip opens the same component key used by the fatigue tables.

4.5.6 Result Viewer

The result viewer on the right renders the computed field on the drum geometry at the most critical time of the selected cycle, which is the timestep carrying the highest stress result. Three tabs switch the field, and Figure 37 shows all three.

[IMAGE: image37.png]

Figure 37. The result viewer, showing a computed field on the drum geometry

• Temperature renders the thermal field in degrees Celsius.
• Displacement renders movement in meters, with a direction dropdown carrying North Direction, West Direction and Vertical Direction, so displacement can be read along a chosen axis. The scale spans negative and positive, so inward and outward movement are distinguished.
• Stress renders Von Mises stress in MPa, running from zero upward.

Note: Each color scale is rescaled to the cycle you have selected. A color in one cycle does not mean the same value in another, so read the scale rather than comparing two cycles by color.

Zoom, pan, rotate and reset work as described in Model Viewers and Color Scales.

4.6 Bulging Inspection

Bulging Inspection answers how far the shell has permanently deformed away from the shape it was built to, and how likely that deformation is to start cracking. Its geometry comes from an inspection campaign rather than from sensors, so this page reports against an inspection dataset instead of a cycle. Figure 38 shows the page.

[IMAGE: image38.png]

Figure 38. The Bulging Inspection page

The headline figure is PSLF: the accumulated plastic strain at a location as a share of the local failure limit set by the fitness-for-service standard. It is recomputed on every processed cycle, with the scanned shape loaded by the operating data of that cycle, and each value falls into a likelihood class, from UNLIKELY to LIKELY, whose bounds your own subject-matter expert sets.

4.6.1 Inspection Data

Inspection Data sets what the whole page reports against. The list carries one entry per turnaround campaign, so the same assessment can be read at each inspection in turn and the change between them compared. A live entry sits alongside them, which takes the geometry from the most recent inspection and re-runs the assessment against the latest cycle.

Info: Selecting a past turnaround shows the condition as that campaign found it. Selecting live shows the latest inspection geometry carried forward to current operation. Neither is a live measurement of the shell, because shell geometry is only measured during an inspection campaign.

A laser scan of a drum this tall can take weeks or months, so a campaign usually covers the locations of most concern rather than the whole shell, and two campaigns may not cover the same ground.

The scan therefore holds still between campaigns, while the PSLF computed on it is recomputed with every processed cycle.

4.6.2 Header Cards

Three cards head the page. Inspection Data, described above, sets what the page reports. The other two summarize the selected dataset. Figure 39 shows all three.

[IMAGE: image39.png]

Figure 39. The three header cards, with the maximum fraction reported once per component group

Max. Plastic Strain Limit Fraction reports the highest PSLF found across every measurement point in that inspection, given once for Shell Body and once for Cone.

Est. Bulging Life reports the estimated years until that maximum reaches 100 percent, projected from a machine-learning prediction of how the measured bulges develop, and names the cycle and date the estimate belongs to.

Note: Reaching 100 percent means the location has reached a condition the local failure check does not accept. It is not a crack, and it is not a failure.

4.6.3 Top Severity Locations

Top Severity Locations ranks the locations carrying the highest plastic strain limit fraction in the selected inspection. Shell Body and Cone are separate tabs, each ranked on its own, and the number of rows in each is set by the Critical Regions field on Bulging Inspection Results. Figure 40 shows the table.

[IMAGE: image40.png]

Figure 40. Top Severity Locations, with the Shell Body and Cone tabs

Likelihood turns the fraction into one of four classes, under the heading Likelihood of Bulging Cracks. Where each class starts and stops is a setting rather than a property of the assessment, and the values below are an example of one configuration:

• UNLIKELY, up to 40 percent
• POSSIBLE, 40 to 60 percent
• PROBABLE, 60 to 80 percent
• LIKELY, 80 to 100 percent

Setting the Likelihood Bounds

A third tab, Likelihood, is where those bounds are set. Upper Bound of Unlikely, Upper Bound of Possible and Upper Bound of Probable are editable fields, Upper Bound of Likely is fixed at 100 percent, and Reset and Save apply what you enter. Figure 41 shows the tab.

[IMAGE: image41.png]

Figure 41. The Likelihood tab, with the class bounds as editable fields

Info: The class bounds are not part of the assessment. They are set by your own subject-matter expert, so the values a build opens with are options rather than a recommendation.

4.6.4 Bulging Inspection Results

Bulging Inspection Results holds two views of the same inspection. Inspection Visualization draws the whole shell as a flat sheet, and Ovality takes a single slice through it. A toggle at the top right switches between them.

Inspection Visualization uses the flat shell map described in Flat Shell Maps, the same sheet the crack pages draw. Figure 42 shows it with the ranked markers in place.

[IMAGE: image42.png]

Figure 42. The unwrapped shell map, with shell courses, plate seams and ranked severity markers

Inspection Visualization carries a Field Selection dropdown with two fields, PSLF and Bulging, and the same map redraws for whichever you pick. Color Min and Color Max set the ends of the color scale, and Apply commits them. A third control, Critical Regions, appears only when PSLF is selected. It sets how many locations are ranked, so the number entered there is the number of rows the Shell Body table carries and the number of markers drawn on the map. Selecting Bulging removes the control. Figure 43 shows the controls.

[IMAGE: image43.png]

Figure 43. The Inspection Visualization controls, with Field Selection, the color scale limits and Critical Regions

Note: The color scale is yours to set. A color on one inspection means the same value on another only if Color Min and Color Max are set the same way on both, so check the limits before comparing two campaigns.

The Bulging Field

What it shows. How far the shell has moved away from the shape it was built to, at every measured point, in meters.

Where it comes from. Each measured point is compared against the design geometry at the same position, and the signed difference is what the color carries.

How to read it.

• Red is outward, away from the axis of the drum.
• Blue is inward.
• Zero is design shape. A drum that had never deformed would be one flat color.
• The ranked markers from Top Severity Locations are drawn on this field as well, labeled Shell 1 onward and Cone 1 onward, so a bulge can be read against the locations the table lists.

The PSLF Field

What it shows. How much of the material’s local failure limit the deformation has consumed, as a percentage on the same flat sheet. The acceptance range is 0 to 100 percent, and a location past the limit reads above 100.

Where it comes from. The PSLF at every measured point. The ranked locations from Top Severity Locations are plotted as markers carrying the same labels, so the table and the map point at the same places.

How to read it. The scale is marked with the four likelihood classes used in the Likelihood column of the table above, so a color can be read straight into a class without going back to the numbers.

Info: The two fields answer different questions. Bulging says how far the shell moved. PSLF says how close that movement has taken the material to its limit. The largest bulge and the highest PSLF can sit at different locations, so a location that stands out on one field may not stand out on the other.

The Ovality View

Ovality, reached from the toggle, shows one cross-section of the drum.

What it shows. Whether the cross-section is still round, at one elevation you choose.

Where it comes from. At each elevation the laser scan gives the measured inner diameter all the way around. Ovality is the spread between the largest and smallest of those diameters, taken relative to the nominal diameter, so a perfectly round section reads zero however large or small it is.

How to read it. The plot reports Radial Deviation: how far the measured wall sits from the design position, in millimeters, at every angle around the circumference. Figure 44 shows the view.

[IMAGE: image44.png]

Figure 44. The Ovality view, with the polar cross-section and the radial scale controls

The blue line is the original profile, a reference at 0.0 mm deviation rather than a measurement.

The red line is the measured wall. Where it sits inside the blue line the deviation is negative and the wall has moved inward; where it sits outside, the deviation is positive and the wall has bulged outward. Hovering reports both values at the angle under the cursor.

Radial Scale sets what the radial axis covers. Min (mm) is the value at the center of the plot, Max (mm) the value at the outer edge, and Step (mm) the spacing between the grid rings. Save applies the setting.

That span decides what is visible. A deviation of a few tens of millimeters across a span of a thousand plots as an almost perfect circle, and narrowing the span to the range the deviation occupies spreads the same data across the plot.

Idea: Set Min a little below the most negative deviation and Max a little above the most positive, with Step giving about five rings. For a trace between roughly minus 80 mm and minus 20 mm, minus 100 to 0 with a step of 20 shows the whole profile.

Note: A round-looking trace is not evidence of a round drum. Check the span set in Radial Scale before reading the shape, and take the Ovality Percentage from the information block rather than from the picture.

The trace is built from one point per degree, so 360 points around the circumference. Irregularity spanning several degrees is measured; anything finer than one degree is not resolved.

Two blocks report the numbers. Selected Elevation Ovality Information gives the elevation you chose from the dropdown with its maximum inner diameter, minimum inner diameter and ovality percentage. Maximum Ovality Information gives the highest ovality percentage anywhere on the vessel and the elevation it occurs at.

Note: The interface applies the ASME requirement that ovality shall not exceed 1 percent, and highlights values at or above 1 percent in orange for further review. Unlike the color bands elsewhere on the dashboard, this one is a code criterion rather than a display convention.

4.7 Crack Inspection

Crack Inspection shows the flaws recorded by one inspection campaign, assessed against the same fitness-for-service method used on Crack Status. It carries the same three widgets as that page, and reads the same way. What differs is what drives it. Figure 45 shows the page.

[IMAGE: image45.png]

Figure 45. The Crack Inspection page

Note: Crack Status, under SPM Monitoring, re-runs its prediction every cycle. Crack Inspection reports the flaws as one turnaround campaign recorded them, so read a figure here as a record of that campaign rather than as the current state.

Inspection Data at the top selects which campaign the page reports. Only campaigns that recorded at least one flaw are listed, so a turnaround that found nothing does not appear here and the list can be shorter than the one on Bulging Inspection for the same drum. The Crack Details table carries the same columns as on Crack Status, with one addition: a Cycle ID column recording the cycle each flaw was assessed against. The FAD Curve and the Crack Specification shell map behave exactly as described in Crack Status.

Because the cut-off is computed for the flaw and the cycle being assessed, the value shown here will not generally match the one on Crack Status, even for the same flaw. That is expected rather than a discrepancy.

4.8 Historical Trends

Historical Trends shows how the way a cycle is run relates to what that cycle costs in fatigue damage, with the whole operating history on one chart, one dot per cycle. Figure 46 shows the page.

[IMAGE: image46.png]

Figure 46. The Historical Trends page

The page holds one widget, the Historical Trend panel: a control row that sets up the view, and the scatter chart beneath it.

4.8.1 Setting Up the View

Select Metric chooses the pairing the chart plots. The available pairing is fatigue damage against cycle length, one dot per cycle.

From and To set the date range, and Apply commits it.

Three inputs place reference marks on the chart. All three are yours to set:

The band isolates the cycles that ran inside the length window you are interested in, so that population can be read apart from the rest. Save fixes the threshold and the band.

Note: Save applies to everyone. The threshold and the band are stored for the dashboard rather than for your session, so changing them changes what every other user of this asset sees.

4.8.2 Reading the Scatter

Each dot is one completed cycle, placed by how long it ran and how much fatigue damage it caused. Hovering reports the Cycle ID, the start and end timestamps, the cycle length and the damage. Double-click a dot and the dashboard opens that cycle on Cycle Inspection, which is where the operation behind it can be examined. Figure 47 shows the chart with its three reference marks.

[IMAGE: image47.png]

Figure 47. The scatter chart, with one dot per cycle and the three reference marks

Cycle Length (hours) runs along the horizontal axis and Damage [%] up the vertical, on a fixed logarithmic scale. Each gridline is ten times the one below, so two dots that look close together low on the plot can differ by a factor of ten.

Three references sit on the chart, and they do not behave alike:

• Average Damage is computed from the cycles in the date range you applied. The slider beneath the axis changes what is on screen without recomputing it.
• High Damage is the threshold you set, and it stays where you put it whatever the range.
• User Defined Optimal Range is the shaded band between the two cycle lengths you set.

The chart shows whether cycle length and damage are related on your drum, isolates the cycles that ran inside a length window, and identifies the low-damage cycles. Cycle length is a proxy for how a cycle was operated, which is why the dots spread vertically at any one length.

Take the Cycle ID of a cycle into Cycle Inspection to see the operating data behind it.

Note: A dot says what a cycle cost, not why it took the time it did. Before comparing two dots, open both on Cycle Inspection and check they are comparable operations.

5. Troubleshooting and Support

5.1 Common Problems and Fixes

Most questions about a number on the dashboard come down to one of the cases below.

5.2 Contact Akselos Support

For anything the articles do not answer, contact Akselos support at [email protected]. That is the route for a dashboard that will not open, a page that is not reporting what you expect, a figure you cannot reconcile against another page, or a request to set up a live data connection.

5.3 Related Articles

This user manual sits in a set with the tutorials and the methodology article, which carry the engineering behind the numbers.

• To follow a fatigue assessment from the operating data through to the estimated life figure, see the tutorial on the fatigue assessment.
• For shell deformation and ovality, how the plastic strain limit fraction is arrived at, and what places a flaw inside or outside the acceptable region, see the tutorial on the bulging and crack assessments.
• To look for the operating pattern that gets the work done for the least damage, see the tutorial on historical trends.

How the structural model is built, divided and configured, and which clause of which standard each step of the analysis draws on, is covered in the methodology article.

If you are still setting up, the article on signing up for an Akselos account covers creating one, and Sign In and Open the Dashboard covers getting access to your organization.

The technical basis behind the numbers on this dashboard is collected in one place.