Michigan’s Capital Preventive Maintenance Program, set up in 1992, published a table crediting a 1.5-inch hot-mix asphalt overlay on flexible pavement with 5 to 10 years of extended service life, a single chip seal with 3 to 6, and crack sealing with up to 3. A footnote under the table governs every row: each range is the life-extending benefit given to the pavement, not the expected longevity of the treatment itself. FHWA reprinted the table in its 2003 Pavement Preservation Compendium.
Set that beside FHWA’s Concrete Pavement Preservation Guide and the footnote starts to matter. The guide gives dowel bar retrofit a treatment life of 10 to 15 years; Michigan credits it with 2 to 3 years of added pavement life and flags even that as needing more data. One figure counts how long the repair holds up, the other how much later the pavement reaches its rehabilitation trigger.
Designing pavement that needs less of this work is covered in long-life pavement design, and the discounting that turns a treatment schedule into a present value is in highway lifecycle cost analysis.
Preservation ends where structure begins
FHWA’s concrete guide reproduces the definitions the agency issued in 2005. Preventive maintenance slows deterioration and keeps or improves functional condition without materially adding structural capacity. Rehabilitation restores structural capacity or adds thickness to carry existing or projected loads. Reconstruction replaces the whole structure and is required when a pavement has failed or become functionally obsolete.
Michigan attached time scales to the tiers: reconstruction designed to last at least 20 years, most rehabilitation 10 to 20, and preventive maintenance adding an average of 5 to 10. Its strategy sent reconstruction to the worst highways, rehabilitation to poor ones, and preventive maintenance to good ones.
The FHWA guide contrasts this with worst-first programming, where pavements deteriorate until the worst one tops the capital list. It calls cost savings the primary benefit of preservation, “but one that has been difficult to substantiate.” Federal condition reporting pushes against these tiers in ways traced in the piece on how agencies rank rebuild candidates.
Right treatment, right road, right time
The phrase appears in FHWA’s concrete guide, Michigan’s program description and MnDOT’s Pavement Preservation Manual. It contains three separate tests.
The right road still has structure worth protecting. Caltrans’ Maintenance Technical Advisory Guide (MTAG, third edition, 2026) says that past a certain point lost strength cannot be remedied by a preservation treatment. MnDOT writes the test as thresholds, recommending a seal coat where traffic is under 10,000 AADT, at least 7 years have passed since the last rehabilitation, wheel-path rutting is 0.5 inch or less, and load-related cracking stays under limits such as 20 feet of alligator cracking in a 500-foot section.
The right treatment matches the distress, and the right time comes from condition data. Michigan prescribes treatments by measured condition rather than by schedule. FHWA’s training guide for local public agencies describes a city that slurry-sealed residential streets in years 5, 12 and 18 and milled and replaced them in year 25; its engineer concluded the network would hold up better if work were prioritized by pavement condition rather than elapsed time.
Caltrans makes the argument at network scale. Every lane mile loses one lane-mile-year of life annually, so a 250-mile, two-lane network loses 500. In the guide’s illustration a worst-first allocation adds back 245, while the same budget spread across cheaper treatments adds 525. The result depends entirely on the life-extension figure entered for each treatment.
Asphalt treatments, thinnest to thickest
Caltrans names oxidation as the primary cause of early asphalt deterioration: binder near the surface hardens and cracks, and water follows. Surface treatments exist to interrupt that sequence, which is also where the damage in freeze-thaw cycles and potholes begins. The mixtures underneath are covered in the site’s reporting on paving materials.
MnDOT considers crack filling practical only where cracking is limited and mostly nonstructural. A fog seal is a light spray of emulsified asphalt, 0.05 to 0.20 gallons per square yard in MnDOT practice, for raveled or porous surfaces and fresh chip seals.
A chip seal (MnDOT’s seal coat) rolls aggregate into sprayed emulsion. It restores friction but does not improve ride, so a cape seal tops it with slurry or microsurfacing. Microsurfacing carries an additive that makes it break chemically, while slurry cures more slowly; both can fill stable ruts. A scrub seal brooms a rejuvenating emulsion into cracks before the chips go down, and MnDOT places it where chip seals no longer work but a mill and fill is too expensive.
MnDOT’s thin hot-mix options run from 0.4-inch bonded wearing courses through 5/8 to 1 inch thinlays to 1 to 2 inch mill-and-overlay work.
| Treatment | What it addresses | What it cannot fix or where it is not used | Published figure (source; what it measures) |
|---|---|---|---|
| Crack seal or fill | Water entry through cracks | Structural cracking; adds no structure | Up to 3 yr (Michigan; added life). Fill 1 to 3 yr, rout and seal 2 to 4 yr (MnDOT; performance period) |
| Fog seal | Raveling, porous surfaces, very fine cracks | Medium or high severity cracks without crack treatment first | 2 to 4 yr (MnDOT; performance period). Crack and fog seals 1 to 4 yr (FHWA local-agency guide; added life) |
| Chip seal | Low-severity cracking, raveling, friction loss, moisture | Ride quality; ruts need separate filling | Single 3 to 6 yr, double 4 to 7 yr (Michigan; added life). 5 to 7 yr (MnDOT; performance period) |
| Slurry seal | Low-severity cracking, raveling, friction, stable ruts | No structure; MnDOT recommends it only under 2,500 ADT | Insufficient data (Michigan). 5 to 7 yr (MnDOT; performance period) |
| Microsurfacing | Same as slurry, plus rut filling and quick reopening | No structure | Single course 3 to 5 yr, multiple 4 to 6 yr (Michigan; added life, more data needed). 5 to 7 yr (MnDOT; performance period) |
| Cape seal | Chip seal benefits with a smoother surface | Larger cracks and potholes must be repaired first | 6 to 8 yr (MnDOT; performance period) |
| Scrub seal | Low and medium severity cracking, raveling, poor friction | Roads over 2,500 ADT in MnDOT guidance | 6 to 7 yr (MnDOT, citing Road Resource.org; performance period) |
| Thin hot-mix overlay | Cracking, raveling, friction loss, roughness | Fatigue or rutting damage in the structure | 1.5 in. on flexible 5 to 10 yr (Michigan; added life). Type A thin overlay 1 to 3 yr high volume, 3 to 5 yr low volume (Caltrans; added life). 1 to 2 in. mill and overlay 8 to 10 yr (MnDOT; performance period) |
Concrete: grinding, resealing, dowels and patches
Diamond grinding removes faulting and studded-tire wear. FHWA’s concrete guide (FHWA-HIF-14-014, 2014) treats cracking in more than about 10 percent of slabs as a sign of a structural problem where grinding may not be appropriate, and warns that faulting redevelops under traffic if load transfer and slab support are not fixed first. Its field record is the deepest of any treatment here. A 1999 study of 76 projects in 9 states found ground pavements went an average of 14 years before a second grind or rehabilitation, with 80 percent lasting at least 11; a 2005 California study put expected longevity near 17 years at 50 percent reliability and 14.5 years at 80 percent.
Dowel bar retrofit sets dowels in slots across joints or cracks to restore load transfer, and belongs only on pavement not already deteriorating structurally. Washington reviewed about 180 lane miles holding roughly 380,000 slots and found distress in under 10 percent after 2 to 14 years.
The guide’s ranges for joint resealing and concrete crack sealing assume failure when 25 percent of the sealant stops working, though it cites silicone sealants observed past 20 years on projects in Arizona and Washington.
Partial-depth repair replaces spalled concrete in the top third of the slab. Done well it can last 15 years or longer; poor materials or workmanship can fail it in 2 to 3.
| Treatment | What it addresses | What it cannot fix | Treatment life (FHWA-HIF-14-014) | Added pavement life (Michigan) |
|---|---|---|---|---|
| Diamond grinding | Faulting, roughness, studded-tire wear, low friction | Structural cracking; faulting returns without load transfer repair | 8 to 15 yr | 3 to 5 yr (more data needed) |
| Joint resealing | Water and incompressibles entering joints | Not applicable to continuously reinforced pavement | 2 to 8 yr | 3 to 5 yr |
| Dowel bar retrofit | Lost load transfer, faulting at joints and cracks | Pavement already failing structurally | 10 to 15 yr | 2 to 3 yr (more data needed) |
| Partial-depth repair | Spalling in the upper slab, scaling | Materials-related distress, dowel lock-up, working cracks, damage at dowel depth | 5 to 15 yr | Up to 5 yr (spall repair) |
Two numbers that measure different things
The FHWA guide’s treatment-life table is adapted from a 2011 SHRP 2 study that also tried to estimate pavement life extension and found insufficient data for most concrete treatments. Treatment life is how long a treatment lasts until the next one is needed. Life extension is the added years of pavement life, a simpler measure whose use for rigid pavements, the guide says, has been very limited.
MnDOT’s performance periods read as treatment life. Caltrans gives both for thin overlays: they last 4 to 6 years but extend pavement life by 1 to 3 years on high-volume roads.
The difference carries into the arithmetic. Caltrans offers equivalent annual cost, the unit cost of a treatment divided by its expected life, as a first screen. Enter treatment life where added life belongs and the cost per year of benefit comes out too low. Anyone running options through a pavement lifecycle cost calculator needs to know which of the two numbers the input is.
What federal-aid funds cover
Michigan’s program history dates federal-aid eligibility for preventive maintenance to the 1991 surface transportation law, ISTEA. Today 23 U.S.C. 116(a) defines preventive maintenance to include pavement preservation programs and activities, meaning a network-level, long-term strategy of cost-effective practices that extend pavement life, improve safety and meet road user expectations. Section 116(e) makes a preventive maintenance activity eligible if the state shows “the activity is a cost-effective means of extending the useful life” of a Federal-aid highway.
FHWA’s operative guidance is a February 25, 2016 memorandum from Walter C. Waidelich Jr., Associate Administrator for Infrastructure, issued to align with the FAST Act. It superseded the 2005 definitions memo and two 2004 eligibility documents:
- Preservation is eligible and encouraged under the National Highway Performance Program and the Surface Transportation Program, because section 1103 of MAP-21 added preservation to the definition of construction in 23 U.S.C. 101.
- Routine maintenance, work done in reaction to an event, a season or overall deterioration that needs recurring attention, is not eligible. Section 116(b) leaves maintaining federally funded projects to the state.
- Preventive maintenance is eligible when the state demonstrates cost-effectiveness to the satisfaction of its FHWA Division Administrator.
- Preventive maintenance eligibility is limited to Federal-aid highways, which exclude roads functionally classified as local roads or rural minor collectors.
The memo names no treatments. Eligibility attaches to the character of the work, and the superseded 2005 definitions, still reproduced in FHWA’s concrete guide, say some routine maintenance may count as preservation depending on timing and the distress. The same crack seal can be a programmed preservation project or a routine task depending on when and why it is applied. The memo also confirms MAP-21 imposed no requirement to run a preservation program.
Where the published figures run out
Several ranges in circulation rest on thin evidence, and the documents say so. Michigan marked its microsurfacing, ultrathin overlay, diamond grinding and dowel bar retrofit figures as needing more information and listed slurry seal on flexible pavement as lacking data. MnDOT attributes its scrub seal figure to Road Resource.org rather than its own records, and its manual covers flexible pavements only, directing rigid pavement questions to its Concrete Office. FHWA’s concrete guide could not report life extension for most of its treatments.
Michigan’s method shows why the added-life column is so sparse. Independent engineers evaluated treatments several years old, surveyed 0.1-mile segments for distress, built performance curves from condition data before and after treatment, and took the life extension as the time the treatment held the pavement above its pre-treatment condition. That requires years of condition records for every treatment on every pavement type, and Michigan’s table has blanks where those records did not yet exist.