Bitumen PG 58-16: Specification, Grade Selection and Export Supply

PG 58-16 is a statement that the binder has been tested and found to satisfy the high-temperature criteria of AASHTO M320 at an average seven-day maximum pavement design temperature of 58 °C, and the low-temperature criteria at a minimum pavement design temperature of −16 °C. Two things are easy to slide past there. Both figures describe the pavement at a stated reliability rather than the air above it. And both have to be earned on a rheometer — they cannot be assigned from a refinery data sheet, carried across from a penetration band, or inferred from the crude slate.

Those two numbers then fix three test temperatures, and the grade name prints none of them. Shear testing runs at 58 °C in two binder conditions, as supplied and after rolling thin film oven ageing. Fatigue is judged at 25 °C on pressure ageing vessel residue, and low-temperature behaviour at −6 °C on that same residue. All three come back out of the grade name with nothing more than a calculator, which means whoever is holding a PG 58-16 certificate can audit it without involving a laboratory at all. The arithmetic is worked through in full further down this page.

A two-sided envelope, not a floor

The most common misreading of a performance grade is to treat it as a minimum standard that any better binder automatically clears. It is not. The high-temperature criteria are a floor on stiffness: the binder must be stiff and elastic enough at 58 °C to resist rutting. The intermediate and low-temperature criteria are ceilings: at 25 °C the aged binder has to stay at or under 5000 kPa on G*·sin δ; at −6 °C it has to stay at or under 300 MPa on creep stiffness; and its m-value must not drop under 0.300. Between them those three cap how stiff, and how unable to relax, the aged binder is permitted to become.

That has a practical consequence buyers rarely draw out. A binder graded PG 64-16 has demonstrated its fatigue behaviour at 28 °C, not at 25 °C, and it has not been shown to clear the 5000 kPa ceiling at the colder intermediate temperature this grade uses. It may well do so; it has not been tested to say so. Under M320 a binder is designated by the highest grade it actually satisfies, so a product correctly labelled PG 58-16 occupies a specific six-degree band at each end — it is not a shorthand for “at least this good”. This is also what stops over-blown, heavily processed or over-aged material from being quietly sold into a low grade: the ceilings catch it.

What the designation leaves out

Performance grading answers one question — across what temperature range has this binder been shown to work — and deliberately answers no others. Crude source, wax content and asphaltene fraction are outside its scope. So is adhesion: nothing in the designation predicts how the binder will bond to one particular aggregate, and moisture damage and stripping fall outside its scope entirely. It is not a mix design. Two cargoes carrying the same PG 58-16 designation can be made from entirely different feedstocks and differ in workability, in ageing behaviour and in how they behave with a marginal aggregate, while both remaining correctly graded. What the designation does supply, and what no penetration or viscosity figure supplies, is a temperature range demonstrated on binder that was artificially aged first.

The high-temperature tests: both at 58 °C

Only one of the three test temperatures is printed on the drum, and it is the straightforward one. The 58 °C in the name is the temperature the dynamic shear rheometer runs at, to AASHTO T315 / ASTM D7175. The grade puts two questions there rather than one. Binder as supplied must show G*/sin δ of at least 1.00 kPa. Binder that has first been through the rolling thin film oven — 85 minutes at 163 °C under continuous air to AASHTO T240 / ASTM D2872, standing in for the journey through a mixing plant — must show at least 2.20 kPa. The limit more than doubles because the two lines are not the same question asked twice. The second concedes that production hardens the binder, then requires the hardened material to keep refusing to rut through the summers that follow. Only the residue line decides anything in practice: a binder reaching 2.20 kPa aged has cleared 1.00 kPa unaged with room to spare, so an offer quoting the unaged figure alone has quoted the number that cannot fail.

The intermediate temperature: 25 °C for this grade

The fatigue criterion sits halfway between the two grade temperatures and is then moved four degrees warmer. Written out for PG 58-16:

• (H + L) ÷ 2 + 4
• (58 + (−16)) ÷ 2 + 4
• 42 ÷ 2 + 4
• 21 + 4 = 25 °C

At that temperature the PAV residue must show G*·sin δ of no more than 5000 kPa. The expression trips readers up twice over. The modulus and the phase angle are multiplied here, where the rutting parameter divides them, so this is a different quantity rather than a rearrangement of the earlier one. And the limit is an upper bound, so a low number is the good result — the opposite reading habit to the two 58 °C lines sitting directly above it on the same certificate. The physics is that an aged binder which is stiff and elastic at once banks the energy of every wheel pass as recoverable strain instead of losing it as heat, and eventually returns that stored energy in the form of a crack.

Because the intermediate temperature is driven by the sum of the two grade numbers, several unrelated grades share it. Any grade whose two halves sum to 42 is tested at 25 °C: PG 52-10, PG 58-16, PG 64-22 and PG 70-28. A fatigue result at 25 °C therefore confirms nothing on its own — it has to be read alongside the DSR and BBR temperatures. What it does do is expose an error immediately: a report offered as PG 58-16 that shows the fatigue parameter at 22 °C was graded against a −22 low grade, and one showing 28 °C was graded against a −10 low grade. Neither is a rounding difference. They are different tests on different assumptions.

The 25 °C collision with the penetration test

PG 58-16 is one of the grades where the M320 intermediate temperature lands on 25 °C, which is also the reference temperature for penetration (ASTM D5), ductility (ASTM D113) and specific gravity (ASTM D70). That matters commercially because this grade is quoted more often than most against penetration-grade alternatives, so both numbers tend to appear in the same email thread.

They are unrelated measurements that happen to share a water bath setting. Penetration presses a needle into unaged binder and reports how far it sank, in tenths of a millimetre. The M320 criterion loads PAV residue — binder taken through the RTFOT and then held 20 hours at 2.1 MPa — in a rheometer, and reports a modulus product in kilopascals. One is a consistency check on the material in the tank; the other is a cracking-resistance check on the material as it will be after years in the road. Neither predicts the other. Where a specification or an enquiry refers to “the 25 °C figure” on a PG binder, name the parameter — G*·sin δ on PAV residue — so the reply cannot come back as a penetration number.

The low-temperature test: −6 °C for a −16 grade

Every bending beam entry on a PG 58-16 report should read −6 °C, which is ten degrees above the −16 °C printed in the grade name. Buyers regularly read that as an inconsistency, or as the laboratory letting the binder off. It is neither. The specification is trading one variable it cannot afford for another it can.

What the instrument does is straightforward. A small beam of PAV-aged binder is supported at each end, loaded in the middle, and two numbers are taken off its deflection curve at the sixty-second mark. Creep stiffness S, capped at 300 MPa, limits how much stress a given amount of thermal contraction can generate — too rigid, and the shrinking binder loads the mix past what the mix can carry. The m-value, the slope of log stiffness against log time at that same sixty seconds, has a floor of 0.300 — whatever stress does build, the binder has to be able to bleed it away. Stiffness sets the size of the problem; the m-value is the escape route.

Then the ten degrees. Bitumen is viscoelastic, meaning its apparent stiffness is fixed by two things at once: how cold it is, and how long the load has been sitting on it. Those two can be exchanged for one another along a shift factor that is measurable, and that exchange is time–temperature superposition. It matters here because thermal cracking is not an instantaneous event. A pavement drifts down towards its annual minimum across the coldest hours of a night, tensile stress accumulating the whole time while the binder works to relax it away. Reproducing that faithfully at −16 °C would mean holding each beam under load for something on the order of two hours. No acceptance laboratory can run a two-hour test per specimen, and across that duration the beam has barely deflected far enough for the instrument to resolve cleanly. The SHRP answer was to spend temperature in order to buy the time back: a 60-second reading taken at L + 10 °C returns the same creep stiffness as roughly two hours of loading at L itself, a shift of a little over two decades along the time axis. So the beam is conditioned to −6 °C and read at sixty seconds. The number is measured at −6 °C; the behaviour it certifies is the binder’s at −16 °C across a cooling period that resembles a real night.

That gives you a one-glance check on a certificate. On this grade −6 °C is the entry you want to see. A BBR result reported at −16 °C means the laboratory ran the test at L instead of L + 10 — the correct condition for a −26 low grade, which does not exist on the M320 ladder — and the result cannot be read against this specification at all.

Which criterion actually binds on this grade — and why that is the problem

With a demanding grade, reading a certificate usually comes down to locating the one line that was close, because that line is where acceptance was decided. PG 58-16 frustrates that habit, and understanding why is the most valuable thing on this page. At 74 °C the envelope is narrow, and it is undemanding at both ends simultaneously: a straight-run binder that gets to 58 °C on the hot side will normally clear 5000 kPa at 25 °C without difficulty and clear 300 MPa at −6 °C by a wide margin. Nominally the RTFOT DSR at 58 °C is the line that binds. It is not a difficult line to cross.

The result is that a passing PG 58-16 report tells you the binder is inside the envelope but says almost nothing about where inside it. That is why the continuous grade matters more on this grade than on any of its neighbours, and it is dealt with in the sourcing section below.

Two exceptions are worth knowing. Waxy and paraffinic residues can still fail the m-value even at a temperature as mild as −6 °C, because wax crystallising in the binder interferes directly with stress relaxation. And heavily processed, air-blown or hot-stored material can fail the 25 °C fatigue ceiling despite passing everything else — which, again, is the ceiling doing its job. If a report shows a failure on this grade, those are the two lines it will almost always be on.

A 58 °C high grade describes a summer that is genuinely warm without being severe: a dark surface reaching an average seven-day maximum of 58 °C is hot to stand on, but it is well short of the 64 °C and 70 °C surfaces recorded across desert interiors and humid tropical coasts. A −16 °C low grade describes winters that produce frost and cold nights rather than a frozen season. Neither half is dramatic. Between them they cover a very large share of the world’s inhabited road network, which is why PG 58-16 behaves as a general-purpose grade rather than a specialised one.

Two mechanisms, one envelope

The moderation can come from altitude or from the ocean, and the difference matters when you get to the selection decision.

Tropical and subtropical highlands. Elevation removes the summer heat while latitude keeps the winter mild. A site at 1,500 to 2,500 metres near the tropics never has a cold season in the continental sense, but it also never produces the surface temperatures its lowland neighbours do. The low side of the grade in these places is genuinely earned rather than nominal: thin, dry air under a clear sky radiates heat away quickly after sunset, so the pavement surface routinely falls well below the air temperature, and it does so fast. Here the binding half of the grade is usually the winter one, and the cooling rate is what makes it bite.

Temperate maritime margins. A large body of water damps both ends of the year. Summers stay moderate because sea breezes and humidity cap the daytime surface temperature; winters stay mild because the same water releases stored heat. Cooling at night is slow, since humidity and cloud act as a thermal blanket and give the binder time to relax whatever stress it builds. In these places the low side of PG 58-16 is often generous, and the half that actually constrains the design is the summer.

Where the grade fits, region by region

• East African highlands. The Kenyan and Ethiopian highlands, Rwanda, Burundi and highland Tanzania — high-altitude cities and the corridors between them, where daytime sun is intense but the elevation keeps the surface maximum moderate and clear nights drive it down sharply.
• Southern African plateau. The Zimbabwean highveld, the Zambian plateau, the Angolan planalto and the higher parts of the South African interior, where winter nights below freezing coexist with summers that never reach desert severity.
• Andean and Mesoamerican intermediate altitudes. Mid-elevation Colombia, Ecuador and Peru, and the Mexican central plateau. Very stable annual temperatures with a large daily range.
• Temperate maritime coasts. Coastal Chile and Uruguay, New Zealand, Tasmania and coastal Victoria, the Atlantic margin of Iberia, and the Black Sea and Caspian littorals of Turkey and Georgia.
• Subtropical uplands of East and Southeast Asia. Yunnan and Guizhou, the highlands of northern Vietnam and Laos, and upland Myanmar — monsoon-moderated summers with cool but frost-limited winters.

These are orientations, not specifications. A performance grade for a real project comes from the site’s own temperature record processed through the road authority’s model at the reliability level that authority uses, and two towns fifty kilometres apart at different elevations can legitimately land on different grades.

Why PG 58-16 appears less often in published tables than the rest of the 58 family

Buyers researching this grade often find fewer references than they expect, and conclude the grade is unusual. It is not; the reference libraries are simply skewed. Most PG 58 demand worldwide sits in cold-winter regions, where the same 58 °C summer requirement is paired with a much colder low grade — PG 58-22, PG 58-28 and PG 58-34 dominate the North American PG 58 market, because a moderate summer there normally comes with a real winter. PG 58-16 is the mild-winter subset of the family, and it occurs where the moderating influence is altitude or ocean rather than latitude. That is a common combination in Africa, Latin America, upland Asia and the southern hemisphere temperate zone, and an uncommon one in the specification libraries most PG documentation was written from.

Where PG 58-16 is the wrong answer

Three misapplications recur. Specifying it for a hot inland or desert site asks a 58 °C binder to survive a 64 °C pavement, and the rutting shows up in the first hot season rather than after years. Specifying it for a low-latitude humid coast makes the same mistake for a different reason: sea-moderated does not mean cool near the equator, and those pavements routinely exceed 58 °C. Specifying it for a continental interior with a real winter leaves the pavement exposed to the transverse thermal cracking a −22 or −28 low grade would have prevented.

There is also a quieter failure mode specific to this grade. Because the summer side sits at 58 °C, it carries less margin against a warming or badly estimated design temperature than any of its neighbours. If your 58 °C figure came from an old dataset, or was derived at 50 % reliability rather than 98 %, re-derive it before committing. Of the two halves of PG 58-16, the high one is where the exposure lies.

The two single-step moves are not mirror images of one another. Stepping to PG 64-16 buys rutting protection and changes nothing about cold cracking. Stepping to PG 58-22 buys thermal cracking protection and changes nothing about rutting. They cost roughly the same in span terms — both take you from a 74 °C interval to an 80 °C one — so the decision cannot be made on difficulty of supply. It has to be made on which failure your site is actually exposed to. The sequence below is the order the evidence should be taken in.

Step 1: settle the summer side first

Start here for two reasons. It is the side on which this grade carries the least margin, and it is the side that traffic can override entirely. Take the average seven-day maximum pavement design temperature from the site’s own record, at the reliability level the road authority applies — 98 % rather than 50 % if you have the choice — and check how old the underlying dataset is. A 58 °C figure derived from a twenty-year-old record deserves a second look before it goes into a contract. If the honest answer is 58 °C, the climatic case for the higher grade is closed and you move to step 2. If it lands between 58 and 64, the specification is telling you to take PG 64-16.

Step 2: apply the traffic adjustment before you compare anything else

Climate is only half of the high-temperature input. Superpave practice raises the high grade for loading as well: one six-degree step where traffic moves slowly, two steps where it stands still, and occasionally a further step on very high volumes. Run that adjustment from PG 58-16 and one step lands on PG 64-16, two steps on PG 70-16.

That produces a collision specific to PG 58-16, and it is the reason so many buyers arrive here undecided. The traffic grade and the next climate grade are the same product. A project whose climate model says 58-16 but which contains bus stops, signalised approaches, a weighbridge, a port gate, a container yard or a long climbing lane is buying PG 64-16 for those areas regardless of what the temperature map says. Two different arguments, one binder.

The practical decision that follows is a supply decision, not an engineering one: whether to run the project on one grade or two. A single grade across the whole job simplifies silo allocation at the plant, avoids cross-contamination between deliveries and removes an entire class of site error, at the cost of buying the higher grade for pavement that does not need it. Two grades cost less in binder and more in handling. Settle it before the grade goes into the contract, because it changes both the product and the shipment plan.

Step 3: settle the winter side on cooling behaviour, not on the minimum alone

A pavement does not crack because a thermometer reached a number. It cracks because tensile stress built up faster than the binder could shed it. That makes the rate at which a site cools every bit as decisive as how far it falls — and it is precisely where the two climates that both produce a PG 58-16 requirement part company.

A maritime site cools slowly. Humidity, cloud and the thermal inertia of nearby water stretch the cooling curve over many hours, and a binder with a modest m-value has time to shed the stress it builds. A highland tropical site under a clear dry sky does the opposite: the surface can lose heat very quickly after sunset, and the same binder has far less time to relax. Two sites recording the same winter minimum can therefore carry materially different cracking risk. Where the record sits marginally between a −16 and a −22 requirement and the site cools fast, the extra step is worth buying. Where it sits marginally between the two but the site cools slowly, PG 58-16 with a healthy m-value margin is usually the better value.

When the answer is the diagonal

If step 1 says the summer was underestimated and step 3 says the winter was too, neither single-step move helps — one of the two failure modes stays uncovered whichever you pick. The honest answer is PG 64-22 at an 86 °C span. It is harder to source, it costs more, and it should be recognised as a different purchase rather than a marked-up version of the same one. A supplier who responds to that requirement by offering PG 64-16 or PG 58-22 with an assurance that it will be fine is offering you one of the two protections you asked for.

What each wrong answer looks like on the road

• Summer underestimated, grade left at 58. Rutting and shoving in the wheelpath, worst on grades, at intersections and wherever traffic is slow. It appears in the first hot season rather than after years, which makes it the most visible and most disputed of the failures.
• Traffic bump skipped. The same failure, but concentrated at stopping, turning and standing points while the open carriageway performs perfectly — a distribution that points straight at the missing adjustment.
• Winter underestimated, grade left at −16. Transverse thermal cracks at fairly regular spacing across the carriageway, followed by water ingress, ravelling at the crack edges and accelerated deterioration of the layers beneath.
• Over-specified where PG 58-16 was correct. Cost and sourcing difficulty with no measurable service benefit. There is also a subtler cost: the binder that reaches −22 from a given source is often softer within its own high-temperature band, so the practical rutting margin inside the grade can be lower even though the grade temperature is identical.
• RAP not accounted for. Reclaimed asphalt brings aged, stiff binder into the mix and shifts the blended grade warmer at the low end. This cuts both ways for PG 58-16: at high reclaimed contents it is frequently chosen deliberately as the virgin grade in a mix designed to land at PG 64-16 after blending, but a design that assumed PG 58-16 in the finished mix and then added reclaimed material has quietly lost its low-temperature protection. Fix the reclaimed percentage and the virgin grade together, never in sequence.

74 °C implies no polymer modification, and should not be priced as if it did

The industry working rule is that a single unmodified binder from a suitable crude will normally hold a useful temperature interval of up to about 92 °C; past that point the two ends pull apart far enough that polymer modification is usually needed to satisfy both at once. PG 58-16 asks for 74 °C. That is roughly three grade steps inside the boundary, and it means the grade should be produced, offered and priced as neat binder. If an offer against this grade arrives described as polymer modified, treat it as a question rather than an upgrade. Either the base material could not span 74 °C on its own — which is information about the feedstock, and not good information — or something the specification never asked for has been substituted into the quotation and priced accordingly.

The same arithmetic leaves generous room for a traffic adjustment. One step up lands on PG 64-16 at an 80 °C span and two steps on PG 70-16 at 86 °C, both still routinely produced without polymer. Only a third step, to PG 76-16 at a 92 °C span, reaches the point where neat capability normally runs out. A project working from PG 58-16 can therefore absorb the heaviest traffic adjustment in ordinary use and still be buying an unmodified product.

The label carries less information here than on any neighbouring grade

This is the warning, and it is the practical consequence of an easy envelope. A binder correctly designated PG 58-16 has a continuous grade — the actual temperature at which each criterion was met, rather than the rounded six-degree step — of at least 58.0 °C but below 64.0 °C on the high side, and at or below −16.0 °C but above −22.0 °C on the low side. That is a wide space to hide in. One cargo might continuous-grade at 58.4 and −16.3, another at 63.5 and −21.6. Both are genuine, correctly labelled PG 58-16. The second carries substantial margin at both ends and will tolerate a hot spell, a batch of reclaimed material or a fast cold night without falling out of grade; the first has almost none, and will also mix and compact at noticeably different temperatures.

On a demanding grade the designation itself tells you a lot, because a binder that reaches it has had to be built for it. On PG 58-16 the designation tells you comparatively little. So ask for the continuous grade, at both ends, in writing. The laboratory already has the figures — they fall out of the same measurements that produced the designation, and stating them adds nothing to the cost of the report. On PG 58-16 they are the only thing separating two offers that are indistinguishable on a specification sheet, and the only way to tell which of the two will still be in grade after a hot spell or a load of reclaimed material.

“Meets PG 58-16” and “is PG 58-16” are different statements

Under M320 a binder is designated by the highest grade it actually satisfies. A certificate reading “meets or exceeds PG 58-16” is therefore worth one question: what does the binder actually grade as? The answer decides whether the phrase means anything, and it does not always point the same way. If the answer is PG 64-22, the claim holds up on the evidence: that binder was shear-tested at 64 °C rather than 58 °C, its fatigue criterion was applied at the same 25 °C this grade uses, and its bending beam criteria were satisfied at −12 °C, a harder condition than −6 °C. Every line covers. If the answer is PG 64-16, it does not follow: the fatigue criterion for that grade is applied at 28 °C, where an aged binder is softer and G*·sin δ is correspondingly lower, so nothing on the report shows the 5000 kPa ceiling being cleared at 25 °C. Overlap between grade envelopes runs in some directions and not in others, and “exceeds” is only true along the axis that was measured. The clean request is a grading report on the tank or batch being loaded, stating the continuous grade at both ends, rather than a conformity statement against a grade the binder may never have been tested to.

What a failure on this grade actually looks like

Given how forgiving the envelope is, an outright rejection at PG 58-16 says something specific about the material, and it is worth knowing what. Almost none of these failures happen on the 58 °C lines. They happen on the two ceilings, and for two different reasons.

The first is relaxation. A paraffinic or waxy residue can drop below the 0.300 m-value at −6 °C while its creep stiffness is still nowhere near 300 MPa, because a crystallised wax network obstructs precisely the flow the m-value measures. −6 °C is a mild temperature for a paving binder, so a source that cannot relax there has a real defect, not a marginal one. The important part for a buyer is that none of the conventional tests will show it: penetration, softening point and viscosity all read normally on a binder whose m-value has failed.

The second is oxidation. Material that has been air-blown, over-processed, or held hot in a tank for too long can exceed the 5000 kPa ceiling at 25 °C while clearing every other line on the sheet, because ageing raises modulus and elasticity together — exactly the combination the fatigue parameter is built to catch. That is the ceiling functioning as designed, and it is the reason a low grade cannot be used as a dumping ground for tired stock. Neither of these two outcomes could have been anticipated from a penetration-grade certificate, which is the whole argument for performance grading.

A penetration grade cannot be converted into a PG grade

No calculation turns 80/100 into PG 58-16, and none turns PG 58-16 into a penetration band. The two systems do not measure comparable things. A penetration figure is one reading at one temperature on binder straight from the tank; a PG designation is a pair of temperature limits demonstrated on binder that was deliberately aged first, at three separate temperatures, on two instruments. There is no bridge between them because the information simply is not present on the penetration side.

What does exist is a loose correlation on the summer end only — softer bands such as 80/100, 85/100 and 120/150 land on a 58 °C high grade more often than on a 64 °C one. That is a tendency in the population, useful for guessing which cargoes are worth testing and worthless as evidence about any one of them. On the winter end there is not even a tendency: the −16 °C requirement is decided by relaxation behaviour after PAV ageing, which penetration does not touch. Two 80/100 cargoes from different crudes can come back as PG 58-16 and PG 58-22, and a third can fail to reach either.

The consequence for procurement is that a grading report describes a quantity of binder, never a product name. A report issued for a tank filled three months ago is evidence about material you are not buying, however honest it was when written. What makes a PG line in a contract enforceable is a DSR and BBR report carrying the same tank or batch reference that appears on the Certificate of Analysis for your own cargo. Where no such report exists, the grade in the offer is a claim rather than a measurement — and a claim and a measurement should not be compared on price as though they were the same thing.

Do not carry a PG 64 plant recipe onto this grade

The 3 Pa·s ceiling at 135 °C, measured to AASHTO T316 / ASTM D4402, is a handling limit and nothing more — it says the binder can be pumped through a terminal and will coat aggregate rather than sit on it. It is not a working temperature. Mixing and compaction temperatures are read off the binder’s own viscosity–temperature line at the conventional equiviscous targets, approximately 0.17 Pa·s for mixing and 0.28 Pa·s for compaction.

On PG 58-16 that has a consequence worth acting on, because this grade is frequently introduced at a plant that has been running PG 64. A PG 58 binder from a given source hits both of those viscosities at lower temperatures than the PG 64 product from the same refinery. Leave the old recipe in place and the plant overheats the binder for no gain: it ages the material in the mixer, spends fuel, and quietly consumes part of the very low-temperature margin the −16 requirement was bought to secure. Ask for the viscosity–temperature data on the binder actually being supplied and set the targets from it. One caveat: the equiviscous method does not hold for polymer modified binders, so if a traffic adjustment moves the order onto a modified grade, the supplier’s recommended mixing and compaction range replaces the calculation entirely.

Ten things to confirm on the certificate before the cargo moves

• A temperature printed beside every rheometer figure. A DSR or BBR value with no test temperature against it cannot be matched to this grade or to any other, and is worth querying rather than interpreting.
• RTFOT, not TFOT. The ageing step ahead of the second DSR must be AASHTO T240 / ASTM D2872. ASTM D1754, the thin film oven test, belongs to the penetration-grade system; the M320 limits were written on rolling thin film residue and do not carry across.
• 2.20 kPa applied to the residue. A recurring error is the 1.00 kPa unaged limit quoted twice — once correctly against original binder and once, wrongly, against the aged result.
• PAV conditioning genuinely performed, to AASHTO R28 / ASTM D6521, 20 hours at 2.1 MPa. Fatigue or bending beam numbers produced from RTFOT residue alone are not M320 results whatever heading sits above them.
• Fatigue at 25 °C. A result at 22 °C came out of a −22 grading and one at 28 °C out of a −10 grading. This single line is the quickest way to catch a report that belongs to a neighbouring grade.
• Bending beam at −6 °C, not at −16 °C, which would mean the laboratory worked at L instead of L + 10.
• The m-value present, not only S. A certificate showing creep stiffness alone has omitted the criterion most likely to have failed.
• Viscosity in Pa·s to AASHTO T316 / ASTM D4402. A kinematic viscosity in centistokes comes from a different instrument measuring a different quantity and cannot be converted to the 3 Pa·s limit.
• Flash point by Cleveland open cup, ASTM D92. A Tag open cup figure is the cutback procedure; finding it on a paving binder certificate means the safety case was built on the wrong test.
• Traceability, and the continuous grade. Tank or batch number, sampling date and test date that tie the report to the cargo you are buying — and, on this grade more than any other, the continuous grade stated at both ends.

If the specification is written to M332 rather than M320

Some authorities have moved to AASHTO M332, and it changes what you have to ask for. The grade picks up a traffic letter — PG 58S-16, PG 58H-16, PG 58V-16, PG 58E-16 — and the RTFOT DSR line is replaced by a multiple stress creep recovery test run to AASHTO T350, which reports non-recoverable creep compliance (J_nr) and percent recovery at 58 °C. Those values are absent from an M320 certificate and there is no way to calculate them from one, so a supplier holding only M320 data cannot answer an M332 specification however good the binder is. There is a second consequence that matters directly to this page: because M332 carries traffic in the letter rather than in a six-degree grade bump, the PG 58-16 / PG 64-16 collision described in step 2 above simply does not arise under that standard — a heavily loaded temperate site stays at 58 on the high side and moves up the letter instead. That is one of the reasons authorities with slow and standing traffic have adopted it. Establish which of the two standards governs your project before ordering, not after. Under either one, the specification that binds is the one written into the sales contract and evidenced by the batch Certificate of Analysis with its supporting test report.