Bitumen PG 64-16: Specification, Climate Fit and Export Supply

Read the name as two design temperatures with a measurement standing behind each. PG 64-16 means the binder has been shown to satisfy the high-temperature criteria of AASHTO M320 at an average seven-day maximum pavement design temperature of 64 °C, and the low-temperature criteria at a minimum pavement design temperature of −16 °C. In a region whose refineries sell almost entirely on penetration, that distinction is the whole reason the grade exists: 60/70 records how far a needle sinks into unaged binder on one bench at one temperature, while PG 64-16 states the two ends of a service window and names the ageing the binder passed through on the way. Neither half of the name is assigned by inspection, deduced from a penetration figure, or inherited from a refinery data sheet.

The grade name prints one test temperature and conceals two more, and all three come back with a calculator — which is why a buyer with no laboratory can still audit a certificate. The shear tests sit at 64 °C, once before ageing and once on rolling thin film oven residue. The fatigue criterion lands at 28 °C, on pressure ageing vessel residue. The bending beam sits at −6 °C, on that same PAV residue. Work those three figures out from the grade name before opening the report, and a certificate graded against a different binder gives itself away in the first column. The arithmetic is set out in full further down this page.

The −16 is the half that gets argued about

Nobody in a Gulf or North African tender questions the 64. The objection, when it comes, is always to the low number: a specification writer asks for PG 64-16 and a supplier replies that a minus sixteen requirement makes no sense in a country where the coast never sees frost. Both parties are usually talking past each other, because they are describing different sites and different quantities.

−16 °C is a pavement surface design temperature at a stated reliability. It is not the average winter night, it is not the coastal winter at all, and it is not an air temperature. It describes the coldest the top of the asphalt is expected to get at an inland site, under a clear sky, at elevation, across the years the design is written to survive. On the interior plateaux of the Arabian Peninsula, the Levant, Iraq, Anatolia and the Maghreb that is an ordinary number rather than an exotic one. The section on climate below explains why.

What the grade does not tell you

What PG 64-16 draws is a climatic envelope, and an envelope is all it draws. The grade says nothing about crude source, wax content or asphaltene fraction. It is not a mix design, it says nothing about how the binder will bond to a particular aggregate, and it does not address moisture damage or stripping. Two cargoes carrying the same PG 64-16 label can come from completely different feedstocks and behave differently in workability, in ageing and in adhesion, while both being correctly graded. What the grade does tell you — and no penetration or viscosity grade tells you — is the temperature range across which the binder has actually been demonstrated to work.

The high-temperature tests: both at 64 °C

Nothing needs calculating on the hot side. The 64 in the grade name is the chamber setting for the dynamic shear rheometer, and two separate criteria are applied there. Before ageing, G*/sin δ on original binder must reach 1.00 kPa. After the rolling thin film oven — 85 minutes at 163 °C under continuous air to AASHTO T240 — the same parameter on RTFOT residue must reach 2.20 kPa. The step up to 2.20 is not the same question asked more severely. Between the storage tank and the paver the binder passes through a hot plant and comes out stiffer, and the second criterion asks whether that stiffened material still refuses to move under a wheel through the summers that follow. In practice only the second one decides anything: a binder that clears 2.20 kPa aged was never in danger of failing 1.00 kPa unaged.

The intermediate temperature: 28 °C for this grade

The fatigue criterion is tested at neither grade temperature. M320 sends it to the midpoint between them and then adds four degrees. Worked out for PG 64-16:

• (H + L) ÷ 2 + 4
• (64 + (−16)) ÷ 2 + 4
• 48 ÷ 2 + 4
• 24 + 4 = 28 °C

At 28 °C the PAV residue must show G*·sin δ of no more than 5000 kPa. Note that this parameter is a product rather than a quotient, and that a low value is wanted. An aged binder that is simultaneously stiff and elastic stores strain energy under each wheel pass instead of dissipating it, and eventually gives that energy back as a crack.

Two practical points follow, and both are specific to this grade. First, 28 °C is one of the warmer intermediate temperatures in the common grade set — PG 64-22 sits at 25 °C and PG 58-22 at 22 °C — and a binder is softer at 28 °C than at 25 °C, so the 5000 kPa ceiling is comparatively easy to clear here. For PG 64-16 the fatigue line is rarely the one that fails. The criteria that actually decide the grade are the RTFOT DSR at 64 °C and the m-value at −6 °C. Second, if a report offered to you as PG 64-16 shows the fatigue parameter measured at 25 °C, the laboratory graded that sample against a −22 grade, not against yours. That is not a rounding difference; it is a different test.

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

The objection raised most often against a PG 64-16 report is that the bending beam was run at −6 °C when the grade name says −16 °C — ten degrees warmer, and apparently a concession. It is not a concession. It is the condition the standard requires, and a laboratory that ran the beam at −16 °C instead would have graded the sample against a different and much colder binder.

The instrument is simple enough: a small beam of PAV-aged binder loaded in three-point bending, with two numbers taken at the sixty-second mark. Creep stiffness S is capped at 300 MPa — the binder must not go so rigid that a contracting pavement generates more stress than the mix around it can carry. The m-value, the slope of the log stiffness against log time curve at that same instant, must reach 0.300 — whatever stress does build, the binder has to be able to let go of it.

The ten degrees are borrowed from time–temperature superposition. Bitumen is viscoelastic: how stiff it appears depends as much on how long a load is held as on how cold it is, and the two are interchangeable along a shift factor that can be measured on real binder. Thermal cracking is not an instantaneous event. An inland pavement radiates heat away for hours after sunset while tensile stress accumulates and the binder works to relax it, so an honest laboratory version at −16 °C would mean holding each specimen under load for about two hours — hopeless as a routine acceptance test, and at that temperature the beam barely deflects far enough to measure reliably. SHRP used the shift to move the test somewhere usable: sixty seconds of loading at L + 10 °C gives the same stiffness as approximately two hours of loading at L. So the beam is conditioned to −6 °C and read at 60 seconds. The reading is taken at −6 °C; what it certifies is how the binder behaves at −16 °C across a real night.

At −6 °C, read the m-value first

Testing this particular grade at only −6 °C has a consequence worth stating plainly, because it changes how a report should be read. At −6 °C most straight-run paving binders are still a long way from 300 MPa, so creep stiffness usually passes with a wide margin and tells you very little. The m-value does not behave that way. Stress relaxation is the first property oxidative ageing attacks, and it is also the property that wax interferes with most directly. On a PG 64-16 certificate the m-value at −6 °C is the number that separates binders that make the grade from binders that do not, and it should be the first low-temperature figure you look at. A report that quotes S at −6 °C and omits the m-value has omitted the decisive result.

Two error patterns are worth naming. If BBR is reported at −16 °C, the laboratory has tested at L instead of at L + 10, which is the correct condition for a different and much colder grade. If there is no BBR result at all, the binder has not been performance graded, whatever appears on the drum, the tank label or the offer.

The climatic case for PG 64-16 rests on one observation: in a dry climate the daily temperature range is far larger than the seasonal one is forgiving. Humidity and cloud act as a thermal blanket. Where they are absent, a surface absorbs solar radiation all day and then radiates it straight back to a clear sky at night with nothing to hold it in. Diurnal swings above 20 °C are routine in the desert interior, and asphalt exaggerates both ends: it is darker than the ground around it, so it runs hotter by day, and it is an efficient radiator, so at night it runs colder than the air above it.

That is the whole grade in a sentence. The same site produces a 64 °C pavement in July and a pavement that can approach −16 °C on a clear January night, and the binder has to survive both without being asked to survive a continental winter it will never see.

Altitude is the variable most tenders overlook

Elevation raises the winter requirement sharply while barely reducing the summer one. Thinner, drier air radiates heat away faster after sunset, so nights are colder; but a dark surface under a high-altitude sky receives more intense insolation, not less, so the seven-day maximum pavement temperature stays high. Inland highland sites in hot countries therefore widen the required span from both ends at once.

The relevant geography is not marginal. Riyadh sits at roughly 600 m and Amman at around 800 m. Large parts of the Saudi interior, the Jordanian and Syrian steppe, northern Iraq, southeastern Anatolia, the Moroccan and Algerian high plateaux and the Atlas foothills combine severe summer insolation with elevation. Every one of them records air temperatures below freezing in some winters. None of them experiences a winter that would justify a −22 grade.

Cooling rate matters as much as the minimum

Low-temperature cracking is not caused by a temperature; it is caused by a stress the binder could not relax quickly enough. That makes the rate of cooling as important as the depth of it. A clear desert night, with no cloud and little water vapour, produces some of the fastest surface cooling rates found anywhere. A binder with a marginal m-value has time to relax stress during a slow maritime cooling and very little time during a fast continental one.

This is why the −16 requirement earns its place in hot countries, and why the failures it prevents look different from northern thermal cracking. Instead of a single catastrophic cracking event on the coldest night of a decade, inland desert pavements accumulate thermal fatigue: many moderate cool-down cycles each winter, each one loading and unloading the binder, gradually opening transverse cracks that then admit water into the structure. The absolute minimum is unremarkable. The number of cycles is not.

Where PG 64-16 fits, region by region

• Arabian interior. Central and northern Saudi Arabia, inland Oman and the interior of the UAE away from the coastal strip. Extreme summer pavement temperatures with winter nights that reach or pass freezing at elevation.
• The Levant and Mesopotamia. The Jordanian plateau, inland Syria, northern and central Iraq. Hot continental summers, genuinely cold winter nights and a wide daily range for most of the year.
• The interior uplands and the lower elevations of eastern Anatolia. These uplands are the textbook case for this grade: intense summer sun at altitude, dry air, and reliable sub-zero winter nights that stop well short of severe cold.
• North Africa away from the coast. Interior Morocco, Algeria and Tunisia, the Atlas foothills, inland Libya and Egypt’s desert routes. The Mediterranean coastal strip is a different climate and usually a different grade.
• Southern United States. The interior south and southwest — Texas, Oklahoma, New Mexico, Arizona and inland southern California — sit in the same climatic band, which is why PG 64-16 appears across that region’s specification landscape while PG 64-22 dominates further north.
• Northern South Asia. The Punjab plains on both sides of the India–Pakistan border, Haryana, Delhi and western Uttar Pradesh, and the Potohar plateau. Summers that punish a binder, and winters mild in absolute terms but cool enough at inland sites to make the low-temperature half of the grade meaningful.

Where PG 64-16 is the wrong answer

Two mistakes are common and they run in opposite directions. Specifying PG 64-16 on the Gulf and Red Sea coast — Dubai, Doha, Kuwait City, Dammam, Muscat, Jeddah — usually buys low-temperature performance the site cannot use. Humidity and the sea moderate the night; the pavement does not approach −16 °C and frequently does not approach −10 °C either. On the coast the honest grades are PG 64-10 or, where traffic is heavy and slow, PG 70-10. Conversely, specifying PG 64-16 where the ground genuinely freezes and stays frozen leaves the pavement exposed to the transverse thermal cracking that PG 64-22 would have prevented, and the saving disappears at the first hard winter. The correct grade is the one the site’s own temperature record produces, not the one the region is assumed to need.

80 °C carries no implication of polymer modification

The working rule across the industry is that a straight-run binder from a suitable crude will hold a useful temperature interval of up to roughly 92 °C, and that stretching both ends beyond about that point normally calls for polymer. PG 64-16 asks for 80 °C, a comfortable twelve degrees inside the limit, so this grade neither implies a modified binder nor justifies being priced as one. When an offer against a PG 64-16 enquiry comes back described as polymer modified, the question to put is which of the two explanations applies: the source cannot reach the grade neat, or the specification is being upsold.

Three grades, one span

PG 64-16, PG 58-22 and PG 70-10 all have a span of 80 °C. They are comparably demanding to manufacture and they are not remotely interchangeable in service. The span tells you how hard the binder is to make; the two grade temperatures tell you where it works. A supplier who offers PG 70-10 against a PG 64-16 enquiry on the grounds that the span is the same has answered a question about production difficulty, not a question about the climate at your site. PG 70-10 is never tested at −6 °C, and the bending beam requirement at −6 °C is exactly what your project is buying.

The margin the grade name hides

The M320 envelope is six degrees wide at each end, so passing tells you less than it appears to. Two binders can both be genuine PG 64-16 and sit in very different places inside it: one might continuous-grade at 67.1 on the high side and −20.4 on the low side, the other at 64.2 and −16.3. Both are correctly labelled. The first carries real margin at both ends and will tolerate a hot spell at the plant, a batch of RAP or a hard night without falling out of grade; the second has almost none. The continuous grade, sometimes called the true grade, is the actual temperature at which each criterion was met rather than the rounded step, and it belongs on the test report. It is the first thing to ask for when two offers look identical on the specification sheet.

Why the m-value is the sourcing question for this grade

The two ends of a PG grade pull against each other. Stiffness and elasticity at 64 °C generally come with a higher asphaltene content and a stiffer maltene phase; compliance and, above all, stress relaxation at −6 °C want the opposite. Air blowing, heavy processing and prolonged hot storage all push a binder up at the top and cost it ground at the bottom. There is no free adjustment, only a trade.

For PG 64-16 the trade lands on a particular property: wax. Paraffinic and waxy residues, which are common across the region this grade serves, frequently reach 64 °C on the high side without difficulty and then struggle with the m-value, because wax crystallising in the binder interferes with stress relaxation precisely where relaxation is being measured. Creep stiffness at −6 °C is rarely the problem; relaxation often is. Heavier, more asphaltic residues generally relax better in the cold but may need blending adjustment to hold the high end. Two refineries can both produce a compliant PG 64-16 from entirely different starting material, and the two products will not behave identically in ageing or workability. That is legitimate variation inside one grade name, not a defect — but it is a reason to grade the batch rather than to trust the label.

Traffic bumping: this grade has room to move

Superpave practice adjusts the high-temperature grade for loading as well as for climate: one step, six degrees, for slow transient loads, and two steps for standing loads, with a further step considered for very high volumes. On this grade that means PG 70-16 under a slow climbing lane or a heavily loaded arterial, and PG 76-16 where traffic stops altogether — bus stops, container yards, toll plazas, signalised approaches. The spans go to 86 °C and 92 °C respectively.

That is a materially more comfortable position than the same exercise starting from PG 64-22, where a single step already reaches 92 °C. Because the low side of this grade is only −16, a one-step traffic bump on PG 64-16 usually stays inside neat-binder territory, and it is the two-step bump that normally moves the order into polymer modified supply. In a region where standing and crawling traffic at ports, border crossings and urban intersections is common, that headroom is a real commercial advantage of the −16 low grade over the −22 one. Resolve the traffic adjustment before the grade goes into the contract, because it changes the product being bought and the price basis with it.

Working temperatures come from the binder’s own viscosity curve

Nothing in M320 tells a plant how hot to run. The 3 Pa·s ceiling at 135 °C is a handling limit — it establishes that the binder can be pumped out of a tank and will coat aggregate — and it is not a mixing temperature. For an unmodified binder the plant temperatures are read off that binder’s own viscosity–temperature chart at the conventional equiviscous targets, near 0.17 Pa·s for mixing and 0.28 Pa·s for compaction. Use the viscosity data for the cargo actually delivered rather than a generic table, because two compliant PG 64-16 binders drawn from different crudes will not share a curve. One caveat matters particularly here: the equiviscous method does not hold for polymer modified binders, so a traffic bump that moves the order to a modified grade replaces the calculation with the supplier’s recommended range.

A penetration grade cannot be converted into a PG grade

There is no arithmetic that turns 60/70 into PG 64-16, and this matters more on this page than on most, because the export trade this grade serves is overwhelmingly a penetration-grade trade. Refineries across the region produce and certify 60/70, 80/100 and 40/50; a PG 64-16 order is usually a penetration-grade production stream that has been performance graded on a specific batch. A needle depth measured at 25 °C on virgin binder says nothing about how the same material shears at 64 °C, nothing about how it relaxes at −6 °C, and nothing whatever about what twenty hours in a pressure ageing vessel will leave behind. The high side of a regional 60/70 does often land somewhere near 64 °C, but that is a tendency rather than a rule, and the low side cannot be inferred at all. Two 60/70 cargoes from different sources can grade out as PG 64-16 and PG 64-10.

The consequence for a buyer is that a PG grading report is batch-specific and does not travel. Production graded last quarter tells you nothing enforceable about the tank being loaded against your contract, and on a penetration-grade production stream the low end in particular can move with the feedstock. Where a specification is written in PG, ask for DSR and BBR results carrying the same tank or batch reference as your Certificate of Analysis. Without them the grade is a description of some binder, somewhere, that nobody has tied to your cargo.