Authors: M. Duval and J. Dukarm
Published in Electrical Insulation Magazine, IEEE
vol.21, no.4, pp.21-27, July-Aug. 2005
doi: 10.1109/MEI.2005.1489986

Abstract: Dissolved gas analysis (DGA) is widely used to detect incipient faults in oil-filled electrical equipment. There is always some degree of inaccuracy in laboratory dissolved-gas measurements, especially at low gas concentrations, which affects gas ratios and other diagnostic calculations. This paper examines the measurement inaccuracy problem in detail using recent CIGRE data. It shows how proper allowance for this inaccuracy can improve transformer condition
assessment and diagnosis by DGA.

Click here to download the paper.



Author: J.J. Dukarm
Published in Electrical and Computer Engineering, 1993.
Canadian Conference on , vol., no., pp.329-332 vol.1, 14-17 Sep 1993
doi: 10.1109/CCECE.1993.332323

Abstract: Dissolved-gas analysis (DGA) is widely used for detection and diagnosis of incipient faults in large oil-filled transformers. Many factors contribute to extreme “noisiness” in the data and make early fault detection and diagnosis difficult. This paper shows how fuzzy logic and neural networks are being used to automate standard DGA methods and improve their usefulness for power transformer fault diagnosis. The use of neural networks for DGA-with or without fuzzy
logic-is discussed, and some related work is described briefly.

Click here to download the paper.



Authors: F. Jakob, P. Noble, J.J. Dukarm, J.J
Published in Power Delivery, IEEE Transactions on
vol.27, no.2, pp.554-559, April 2012
doi: 10.1109/TPWRD.2011.2175950

Abstract: Dissolved gas analysis (DGA) has been used to classify the type and severity of faults in transformers. The method commonly used to identify severity is to measure the total fault gas concentration and its rate of change, without regard to the relative concentrations of individual gases. Thermodynamic analysis indicates that the energy
required to form gases increases in the order CH4 <; C2H6 ≤ CO ≤ C2H4 <; H2 ≪ C2H2. Based on these results, an energy-weighted dissolved gas analysis (EWDGA) is proposed, where concentrations of individual gases are multiplied by a weighting factor that is proportional to the energy required to produce each gas. Statistical calculations are used
to show that EWDGA is more sensitive to high-energy faults than simple DGA, hence minimizing the damage caused by those faults.

Click here to download the paper.



Authors:  F. Jakob and J.J. Dukarm
Published in Power Delivery, IEEE Transactions
vol.30, no.4, pp.1941-1948, Aug. 2015
doi: 10.1109/TPWRD.2015.2415767

Abstract: Conventional practice for transformer dissolved gas analysis (DGA) is to use concentrations of several fault gases, with or without total dissolved combustible gas, for evaluating apparent fault severity. We suggest a simpler approach based on the normalized energy intensity (NEI), a quantity related directly to fault energy dissipated within the transformer. DGA fault severity scoring based on NEI is shown to be sensitive to all IEC fault types and to be more responsive to shifts in the relative concentrations of the fault gases than scoring based on fault gas concentrations. Instead of eight or more gas concentration limits, NEI scoring requires only two or three limits that can be empirically derived to suit local requirements for any population of mineral-oil-filled power transformers.

This is an open access paper — available from IEEE Xplore via free download. Click here.



TOA4’s gas analysis report contains several alphabet-soup items which are explained here along with a couple of related items. In each case we start with the text label you see on the page or report, then the database field name, then the explanation. Most of this is general information about dissolved-gas analysis (DGA) in power transformers and other oil-filled apparatus, so even if you don’t use Transformer Oil Analyst (TOA) software, this may be of interest.

But first a word about units. Dissolved-gas concentrations in insulating oil are usually expressed as parts per million by volume (ppm), which is the same as microliters of gas per liter of oil. Gas-in-gas concentrations are usually expressed as percent by volume. One percent by volume is ten thousand ppm. All of these volumes are adjusted to standard temperature and pressure (101.325 kPa and 273 K), informally one atmosphere at zero Celsius.

A frequently used term — combustible gas — should be explained before we go on. Combustible gas is gas that is, well, combustible — it can burn if mixed with air in the right concentration. In the context of DGA, the combustible gases are hydrogen (H2), methane (CH4), ethane (C2H6), ethylene (C2H4), acetylene (C2H2), and carbon monoxide (CO).

Total gas (totalgas): This is the total dissolved gas content of the oil, often expressed as percent by volume to make the numbers smaller. Total gas is typically about ten percent or less. Total gas is measured – it is derived from the total area of all the gas peaks in the chromatogram. I am not sure what this number is good for, but maybe someone will explain.

TDCG (tdcg): Total Dissolved Combustible Gas (ppm) is calculated by adding up the concentrations of the dissolved combustible gases (see above). Usually a very large part of TDCG is carbon monoxide, so its value for fault detection and diagnosis is not great.

TCG (tcg): Total Combustible Gas is the percent by volume of combustible gas in the gas found in the headspace or in the gas space in a gas relay.

Equivalent TCG (etcg): Equivalent Total Combustible Gas (ETCG) is an estimate, calculated from the dissolved-gas concentrations and the oil sample temperature, of what the TCG would be in a gas space in equilibrium with the oil. If you are thinking of pumping or handling the oil, ETCG tells you what the TCG would be in air spaces near the oil. If you have a gas space sample and an oil sample that were taken at the same time, the TCG of the headspace gas should be about the same as the ETCG of the oil. If ETCG is much larger than the TCG, it might mean that combustible gas is being formed in the oil faster than it is diffusing into the gas space. If ETCG is much smaller than TCG, it might mean that some gas was formed in the oil so suddenly that it went up into the gas space in bubbles faster than it could dissolve in the oil. That is what I have been told. Of course, in order to interpret ETCG and TCG in that way you have to be very confident that the numbers are correct.

Estimated safe handling limit (eshl): Estimated Safe Handling Limit (ESHL) is an estimate of the lower flammability limit (LFL) of the combustible gas mixture which the dissolved gas in the oil would give off into air in contact with the oil. Expressed in percent like the ETCG, it is calculated from the dissolved-gas concentrations, the oil sample temperature, and the lower flammability limits of the individual combustible gases. Low values of ESHL are “worse” than high ones. If ESHL is less than ETCG, there might be a fire or explosion hazard if the oil with its existing combustible gas content is exposed to air while being handled or pumped. It is important to understand that both ESHL and ETCG are rough approximations, so very fine consideration of second and third decimal places is not appropriate when comparing them. Also note that ESHL pertains only to mixtures of gas in air, not to headspace gas in sealed equipment, which typically has very small oxygen content.

Total heat gas (thg): Total Heat Gas (THG) is the sum of the concentrations of methane, ethane, and ethylene. These three gases are formed by mineral oil at high temperature. Below 300 C methane tends to accumulate fastest, and above 700 C ethylene is dominant. Ethane is in the middle. Roughly speaking, if the concentrations (from highest to lowest) are in the order methane, ethane, ethylene, then the fault temperature is low. If the order is reversed, the fault temperature is high. If ethane is dominant, the fault temperature is medium. Total heat gas is used in some diagnostic gas ratios, especially THG/acetylene, which compares “heat gas” to “arcing gas.”

Total partial pressure (totalpartpress): Each gas dissolved in a fluid such as transformer oil has a pressure (its partial pressure, expressed in atmospheres) which is determined by the number of molecules of the gas, the fluid volume, and the fluid temperature, as though the fluid were simply an empty container. If we add up the respective partial pressures of the combustible gases, plus carbon dioxide, oxygen, and nitrogen, we get the Total Partial Pressure. If the total partial pressure is close to the ambient gas pressure in the tank, the oil is close to saturation and could have a tendency to form gas bubbles in case of a fault, which would add more gas to the oil.



 One of the download links at the bottom of the equipment list page provides a DGA Summary report in .csv format. To use the summary report, first filter and sort the equipment list to show the equipment you want to include in the report. Then click the “Export DGA report data” link to download the report. Like all .csv files, it can be viewed and edited in a spreadsheet.

For each equipment item contained in the equipment list as filtered and sorted, the report contains a row of data consisting of basic equipment information, the latest DGA sample date and re-sample date, gas concentrations, and the DGA condition code, diagnosis, and “chinese summary”.

So far, we are aware of two different uses for this report, described below.

Example 1: Problem report for maintenance personnel

First, be sure that the “In-service items only” filter is checked so that the top of the report will not be occupied by retired or failed equipment. Filter the equipment by Owner or Region, depending on how your equipment maintenance is organized, then set Status = “Abnormal” and Sort = “Substation” before downloading the DGA Status report.

Example 2: Watch list for critical equipment

The summary report can be used to keep track of latest test results for equipment which is being tested frequently – for example, transformers which are known to be under unusual stress during the summer. To list exactly the right equipment, edit each of the equipment items you want to track and insert a text “tag” expression such as “OVERLOAD WATCH” into the Equipment Remarks (eqp_remarks) field. After each batch of new test results is uploaded into TOA4 Online, type the tag expression into the Keywords filter to display your list, then download the status report.


The S. D. Myers Inc. laboratory reports the ratio of interfacial tension (IFT) and acid number as an indicator of the extent of oxidation of insulating oil. This ratio is the inspiration for TOA4’s Fluid Quality Index, which is calculated with an extra factor of 1000 to ensure that typical values of the index are not tiny decimal fractions. Here is the formula:

fqindex = 1000*acidnum/ift

Here, acidnum is the acid number (mg KOH/g), and ift is the interfacial tension (mN/m). The fqindex is calculated so that low values are good and high values are bad. Clean new oil has an fqindex below 1.0, and when fqindex reaches 10, the oil is in poor condition.

While fqindex is good for trending the approximate oil quality and for detecting sudden changes which may need investigation, the acidity and IFT values provide a direct basis for triggering maintenance activity such as oil treatment.

For in-service mineral transformer oil, the acid number starts at about 0.01 mg KOH/g and is considered poor when it has increased to 0.10 or 0.20; at 0.40 it is completely unacceptable because sludge formation begins at about that level of acidity. The IFT starts at about 50 mN/m and declines as oil oxidation produces polar byproducts. At an IFT of about 25 or 30 mN/m, the oil should be reclaimed, and at about 22 mN/m sludge formation begins. How can we use this information to derive norms for fqindex?

Observe that if acid number is less than 0.10 and IFT is greater than 30, then fqindex is less than 3.33; that is, if fqindex exceeds 3.33, then either acid number or IFT must be outside the acceptable range. Using the more relaxed limits of 0.20 for acid number and 25 for IFT, the corresponding limit for fqindex is 8.00. On the other hand, if acid number is greater than 0.40 and IFT is less than 22, then fqindex is greater than 18 and sludge formation has probably started.

The arithmetic exercise above suggests the following set of critical values for fqindex:

Warning level FQ index Meaning
2 (alert) 3.3 Consider oil reclamation
3 (warning) 8.0 Oil reclamation needed
4 (alarm) 18.0 Sludge formation assured


What is the difference between kv_ratings and ratedkv in TOA4?

The ratedkv field — a “legacy” field inherited from TOA3 — is numeric and designates the rated kV of a piece of equipment, or the highest rated winding kV of a transformer. The kv_ratings field is text and can be used to display multiple winding kV’s, usually separated by slashes; it was introduced in TOA4 as a companion to mva_ratings, which is also text.

For analysis purposes, does it matter which of the two fields we use?

You can use either, both, or neither, as you like. Other than the fluidtype, equipnum/serialnum, and apprtype, all the equipment information that the analysis needs is implicit in the analysis norms that you assign to the equipment. Over the years we have discovered that the equipment “nameplate” information given in the TOA database is often either missing or wrong, so we have learned not to trust it for analysis. Whatever equipment information is present is there for the convenience of the test results reviewer, for finding things in the equipment list, for display on reports, and so on.


In several views of the TOA4 equipment list, and also in some reports, previous and present condition codes for DGA, fluid quality, and moisture-in-oil are displayed together to make changes easy to see. For example, 1/1 means that the code was 1 previously and is also 1 as of the latest sample. And 1/3 would mean that the code was previously 1, but the condition has apparently worsened to 3 as of the latest sample.


 To organize TOA4’s equipment list so that items which are due or almost due for DGA sampling are at the top, choose the “DGA status” columns and then sort by “Next DGA.” All of the equipment items which have a blank Next DGA date go to the bottom of the list.

Similarly, choose the “FQ status” columns and then sort by “Next FQ” to organize the equipment list to show priority for fluid quality sampling.


 Each equipment item has an “In service” checkbox – visible when editing the item – which should be checked if the equipment is in service and unchecked if the equipment is not in service, i.e., retired or permanently de-activated. In an equipment import or export file, this field is called “in_service” and has a value of 1 (true) or 0 (false). The default value, if no value for in_service is specified when the equipment is imported for the first time, is 1.

The “In-service items only” checkbox below the Apparatus Type filter at the top of the equipment list page activates a filter which excludes all equipment items for which in_service is false. This prevents, for example, all the failed equipment from occupying the top of the list when you are sorting by Assessment or Next DGA.

The “In-service items only” filter is different from the other filters in that it is not turned off by the Reset button. To deselect this filter, you must manually un-check the checkbox.


 If you are a TOA4 Online user and have a web-enabled cell phone, Blackberry, or PDA, you might like to try out TOA4’s new mobile access feature. Point your phone’s web browser to the TOA4 Online address, but add “/m” at the end of that address.

Some cell phone companies aren’t very good with handling “https”, so if your phone complains about the security certificate, try using “http” instead.

You still have to log in using the same ID and password that you normally use for TOA4 Online.

If you use the mobile access feature, we would appreciate some feedback on how it treats you and what we should do to make it more useful for you.

Here are some example pages. The “home” page (not shown) is an abbreviated equipment list, which is always filtered to keep the list short. Items with an abnormal conditon are red. If you click on an equipment item in the list, its detail page is displayed. The bottom of the equipment detail page (not shown) contains a summary of analysis results, with abnormal items in red.

Equipment Detail












If you click on (say) the “DGA” link in the results summary, a report showing the latest DGA results (with abnormal items in red) is displayed. Individual data items are linked to corresponding graphs, and the DGA diagnosis (if the result code is greater than 1) is linked to a Duval triangle.

DGA Report (bottom)

Triangle Chart













Early one Monday morning a TOA4 Online user reported, not too cheerfully, having “accidentally” deleted all of the equipment (and therefore all of the test data too) from her company’s TOA4 database. Anything we could do? at all? I could hear her boss’s heavy breathing in the background.

Yes, of course we could fix it, but I had to go find Farmer Neil, our server administrator, who was far out on the land with his dynamite, air compressor, and jackhammer preparing the still-frozen soil for this year’s planting of canola (hey, this is Canada). Pushing the hood of his parka back and mopping his forehead with a large red hankie, Neil just said, “Yep, we’ll get her all fixed up in no time. Have to work my way closer to the farm house, though, because my wireless isn’t too good this far out, eh?”

Deleting data does not actually remove anything from TOA4’s database — each deleted item is simply marked as “deleted,” and from then on the software pretends that the item is not there. Every few days the database is “compacted” to remove all the deleted items.

Since the user’s deletion event was recent, and we knew the user’s login ID, it was easy for us to pinpoint in TOA4’s transaction log the exact moment when the mass deletion took place. A simple script was uploaded to the TOA4 Online server from Farmer Neil’s grimy old laptop in the warmth of the big red tractor’s cab. The script rummaged through the database and un-deleted all of that customer’s equipment and test data that had been deleted at that particular time.

Usually these things happen late on Friday afternoon when users are worn out and hurrying to finish something so they can go home for the weekend. This was a new case — a user who was bleary-eyed from the weekend and started working before she had her Monday morning coffee!

The moral of this story: Don’t click any delete button (or the OK button confirming the deletion) until after you have had your coffee. If you do, contact Delta-X Research right away, and we will fix it, as soon as we can find Farmer Neil and get him within wireless range.


Some older standards and technical publicationsuse dynes/cm for IFT units, and some labs apparently still use those units in their reports. Several years ago the use of CGS (centimeters-grams-seconds) units in science and engineering was deprecated. Since then, mN/m (millinewtons per meter) has been used for IFT, and that is what TOA3 and TOA4 use. Since numerically mN/m is equivalent to dynes/cm, no data conversion is necessary for switching to the newer units.

Beware that some labs and publications use N/m (newtons per meter) for IFT. The difference is easy to recognize – IFT values in mN/m are typically in the range 10-50, but the values are a thousandth as large (0.010-0.050) if they are expressed in N/m.


Importing equipment information into TOA4 Online from TOA3, or from TOA3-friendly data files, is easy.

  1. If transferring equipment information from TOA3 to TOA4 Online, export the equipment information from TOA3. To do this, click TOA3’s Equipment button to view the equipment list. Then choose Export in TOA3’s File menu. Specify a file name (and remember which folder the file will be in) and click OK. Select the “entire database table” option in the export options dialog, then click OK to do the export.
  2. Open the equipment data file in a spreadsheet such as Excel.
  3. This is a good chance to clean up any visible problems with EQUIPNUM, SERIALNUM, APPRTYPE, LOCATION, or other important columns.
  4. If FLUIDTYPE is blank for any of the equipment items, fill in OIL for mineral oil, ESTER for ester-type fluids, LFH for other hydrocarbon fluids, and SIL for silicone.
  5. Delete the following columns, which are not needed for TOA4 — remove the columns, don’t just leave them blank. EQUIPCOND, RULES, DGRESULT, NEXTDG, DGREMARK, FLUIDCOND, PCBSTATUS, FQRESULT, NEXTFQ, FQREMARK, RECDATE.
  6. Use Find and Replace to replace commas and tab characters found in any of the fields with semicolon (;). Also make sure that there are no carriage returns or line breaks in any of the fields.
  7. Sort the spreadsheet by FLUIDTYPE and APPRTYPE and RATEDKV.
  8. Create a new column with title “norm_name”.
  9. Fill in the norm_name column with the names of appropriate TOA4 analysis norms. It may be useful to print or download TOA4’s list of analysis norms for reference, to avoid errors. With the spreadsheet sorted as suggested above, it is usually possible to fill in a norm_name for one equipment item and use the spreadsheet’s copy-down feature to fill in the same name for all similar equipment.
  10. Use Save As… to save the spreadsheet as type CSV (comma delimited), with a filename ending with the suffix .csv.


When you are trying to import test data into your TOA4 Online account, there can be various reasons why either TOA4 refuses to accept the data file or some of the data records don’t get imported. Here are some cases that users have encountered so far:

File upload fails

  1. Error parsing data file. Usually this means that you have clicked the wrong thing and tried to upload a spreadsheet (.xls) or word processing (.doc) file. Sometimes it means that the file has been corrupted and contains non-text gibberish.

Import operation fails

  1. Missing or mis-named required columns. For example, if the apparatus type (apprtype) column is not found, neither equipment nor test data can be imported. Often it turns out that the column is not missing, but its name is spelled wrong.
  2. Missing sampledate column. This one is mentioned separately because it can happen in an unexpected way. If the data file has a sampledate column, but you do not designate the correct date format when uploading the file, all the sample dates will be treated as blanks, and the sampledate column will appear to be missing from the uploaded file.
  3. Duplicate columns. There are two columns which either have the same name or which have different names that TOA4 considers to be equivalent (e.g. location and substn_name).

Some data records are imported, but others are not

  1. When you are importing equipment, and some of the equipment numbers or serial numbers (in combination with the apparatus type) already belong to equipment in the database, those records are not imported.
  2. When you are importing data, and some of the equipment numbers or serial numbers (in combination with the apparatus type) do not belong to any of the equipment in the database, those records are not imported.
  3. When you are importing data, and the Confirm Operation dialog says that some tanks need to be created, the data records that need the new tank(s) are not imported unless you check the Create new tanks? checkbox to give permission for creating the new tanks.


Importing test data in TOA4

When you import new test data in TOA4, the dialog that appears just after you click “Import test data” contains a checkbox that tells TOA4 to do an analysis immediately upon completing the import. Normally, this checkbox should be left “checked” unless you are importing many thousands of samples at once. Really huge import files should be broken up into smaller ones anyway.

If you do import new test data without doing an analysis, an “Analyze new” button should be displayed below the equipment list. Be sure to click that button to calculate analysis results before starting to review UNREVIEWED data. After all new data records have been analyzed, the “Analyze new” button is no longer displayed. Note that if the equipment “owning” the data does not have a correct norm name associated with it, its data will remain unanalyzed until the norm name problem is fixed.

If the equipment list is filtered (for example, restricted to a particular substation or apparatus type) when the “Analyze new” button is clicked, the analysis is done only for the equipment selected by the filter. When you are setting up a new database and have thousands of records of unanalyzed data, it is a good idea to use this feature to run the analysis in a few smaller batches rather than trying to do all of the data at once.

In some cases, if you run an analysis for tens of thousands of new data records all at once, the TOA4 Online web server may “time out” and report an error while waiting for the analysis to be finished. This does not mean that the analysis has failed – the web server has simply given up on waiting for it to finish, but the analysis does get done anyway.

Importing test data in TOA3

After you import GAS and/or FLUID data in TOA3, remember to run a batch analysis to process all the new data. In the File menu, go to Batch Analysis.

Note that the batch analysis willl skip all samples not having a valid name in the GASSTD and/or FLUIDSTD fields, matching standards that are in your TOA3 database.



  1. The first row of the file must contain field names recognized by TOA3. Order is not important, and extra fields not recognized by TOA3 are allowed but ignored.
  2. Items in each row must be separated by either commas or tab characters (ASCII character #9). Only one kind of separator character (comma or tab) should be used – not both in the same file. Text items should be enclosed in quotes (“).
  3. Nonblank values must be given for the fields EQUIPNUM, APPRTYPE, TANK, SAMPLEDATE, and SAMPLENUM. The value of SAMPLENUM can be anything. Its only purpose is to discriminate between records which have identical values for the other four identifying fields.Note: The equipment identification (EQUIPNUM + APPRTYPE) must agree exactly with the identification used in the TOA3 database, or the data will not be visible to the user after importing into the database.
  4. If dissolved-gas data and fluid test data for the same oil sample are provided in separate data files, the TANK and SAMPLEDATE values given in each file must be exactly the same.
  5. Nonblank values must be given for GASSTD and/or FLUIDSTD, identifying the relevant analysis standards, which must also be present in the TOA3 database.
  6. Text data items must not contain commas, tab characters, carriage return characters, embedded quotation marks, or binary data.
  7. Numeric data items must not contain units or other non-numeric characters. For example, 37 C is not a valid entry for FLUIDTEMPC because the letter C should not be included. Likewise, <5 is not a valid entry for H2 because the less-than symbol is not allowed.
  8. Blank data items must be completely empty, not “” or ” “.


  1. Unused columns can be omitted from the file. For example, if the lab does not report values for propane and propylene, then the C3H8 and C3H6 columns can be omitted from the data file. Similarly, if trace elements are not reported in a set of fluid quality results, then the corresponding columns can be omitted from the data file.
  2. To insure compatibility with the client’s TOA3 database, the client should provide the EQUIPNUM + APPRTYPE + TANK + SAMPLEDATE + SAMPLENUM identification for each sample submitted to the lab.
  3. Furthermore, for each sample submitted the client should also provide an analysis standard ID to be included in the sample data record GASSTD column (for dissolved gas samples) or the FLUIDSTD column (for fluid quality test samples) or both (for files containing both DGA and fluid test data).Note: If a standard ID is not included in the data record, the Transformer Oil Analyst user will have to insert one manually before TOA3 can generate analysis reports based on the imported data.
  4. Please use four-digit years in date values! Unless the client specifically requests otherwise, date values (e.g. for SAMPLEDATE) should be given in ISO 8601 format.Acceptable variants are:yyyy-mm-dd oryyyy/mm/dd oryyyymmdd.For example, these date expressions fit the recommended format:2006-06-13 or2006/06/13 or2006.06.13 or20060613.Note: If dates are given in a different format (e.g. mm/dd/yy), then they are recognized and interpreted correctly only if the importing PC’s Windows Short Date format is compatible with the date format used in the data file. Furthermore, if two-digit years are given, TOA3 will interpret them as belonging to the current century. For example, 99-06-13 or 6/13/99 would be interpreted as June 13, 2099. This approach to interpreting two-digit year data is mandated by the ISO 8601 standard.
  5. TOA3 expects import files to have an extension of .TXT or .CSV. Please be careful in naming files because the client may “clobber” other data files or have other problems if he has to rename the files he receives from the lab. One suggestion would be to include the test data type, month, and day in the filename, e.g. GAS0713.TXT or FQ0713.TXT.


The GAS and FLUID data files exported from TOA3 are importable into TOA4, but “to guarantee best results” it is a good idea to inspect the files and modify them, if necessary, before importing into TOA4. The same considerations apply when importing from TOA3-friendly data files prepared by your lab. You can review and edit these data files in a spreadsheet program.

  1. Make sure that the equipment identification (EQUIPNUM+APPRTYPE) in the data file agrees with the corresponding values in TOA4.
  2. Make sure that the TANK values used in the data file agree with the ones you are using or want to use in TOA4. Watch out especially for TANK mismatches between GAS samples and FLUID samples and correct them.
  3. If at all possible, use the ISO (yyyy-mm-dd) date format in all your data files. If you don’t, then when uploading to TOA4 you must designate which date format is used.
  4. Unless the times are really significant, delete the time portion (usually “00:00:00″) from all the dates. If you don’t, all the dates displayed in your TOA4 reports will show the time too.
  5. Ordinarily it is best to delete the SAMPLENUM column. If you have been putting syringe numbers in the SAMPLENUM field, then re-name that column to CONTAINER_ID. Beware that if a FLUID record and a GAS record with the same sample date have a different container ID, then they will be treated as distinct samples by TOA4.
  6. It is good practice, but not strictly necessary, to delete all the unneeded data columns from the file before uploading to TOA4. Actually remove the whole column, don’t just clear it and leave a vertical gap. These unneeded columns would be:
    • All the columns that contain no values.
  7. If the COMMENT field contains tabs or commas, replace them with something else, such as dashes (-) or semicolons (;).
  8. If the data file came from a lab, eliminate any non-numeric characters from values that are supposed to be numeric. For example, change “< 1″ or “< 5″ or “ND” or “Trace” or anything else meaning “no measurable quantity detected” to “0” (zero). That really should be zero, not the detection limit.
  9. The KVD1816 field in a TOA3 data file is assumed by TOA4 to contain D1816 dielectric breakdown values measured with a 2-mm gap. If the KVD1816 values in your data file are for a 1-mm gap, change the column name to “d1816_1″. If you have some values for 1-mm gap and some for 2-mm, then you must put the 1-mm values in a column named “d1816_1″ and the 2-mm values in a column named “d1816_2″. If you are not sure what gap size was used for the D1816 measurements in your data file, you can usually guess because the average for 2-mm gap measurements is normally greater than 35 kV and often greater than 45 kV.

When in doubt about the proper name of a data field or the units associated with it, go to TOA4 Help and look under “Reference information” for the type of data that you are dealing with.