How do I tell a mixing problem from a rolling problem?

Scratch through the film in a light band and a dark band, and look at what the difference does through the thickness. If the shade difference runs all the way down to the substrate, the material that landed there was already a different shade from the material next to it. That is pigment distribution, and it happened in the bucket. If the film is the same shade through its depth and what differs is how much light comes back off the surface, you are looking at sheen rather than color, and the cause is application. Lap marks and roller banding do that, and they have their own page. The second test is geometry. Mixing faults follow the pour: they change where one kit ended and the next began, and they swirl rather than run straight. Application faults follow the applicator: they run parallel, at roller width, and line up with where passes started and stopped.

What actually causes the pigment to be unevenly distributed?

Four things, in roughly the order they turn up on real jobs. Under-mixing is first and by a distance. Pigment is heavy and it does not disperse on its own. A minute of stirring produces material that looks uniform in a bucket and is not, because the color is carried in a paste that needs shear to break up, not just motion. Settled pigment in storage is second. A kit that has stood for months has a denser, more heavily pigmented layer at the bottom. Pouring off the top and mixing what comes out gives you a pale batch and, later, a dark one. Scraping the sides and bottom is third. Unmixed resin clinging to the container wall is both off-ratio and off-color, and it usually goes onto the floor in the last few square feet of the batch, which is why the end of a pour streaks more often than the middle. Batch-to-batch inconsistency is fourth: two kits mixed for different lengths of time, at different temperatures, or by different people, are two different colors even when they came off the same production run.

How long should I actually be mixing?

Longer than feels necessary, and by the clock rather than by eye, because the whole problem is that under-mixed material looks mixed. Most manufacturers specify a mixing time and a mixing speed on the data sheet, and for a pigmented system it is usually measured in minutes rather than seconds. Read that number and set a timer, because the human tendency is to stop when the swirls disappear, which happens well before the pigment is dispersed. Use a mixing paddle on a drill at a controlled speed rather than a stick. The goal is shear, not stirring, and a paddle that folds material from the bottom back through the middle is doing something a stick cannot. Scrape the sides and the bottom partway through and then keep mixing. A batch that was mixed perfectly for three minutes and then had unmixed material scraped into it at the end is an unmixed batch. And do not mix so fast that you whip air in. That trades a color problem for a bubble problem, which has its own page.

Can I fix streaking without recoating the whole floor?

Rarely, and it is worth being straight about why, because the intuitive repairs all make it more obvious. Spot-coating a pale area with a fresh batch puts a third shade on the floor. Even a correctly mixed batch of the same product will not match a badly mixed one, so the patch reads as a patch. Buffing or polishing changes sheen, not color. If the difference runs through the film, there is nothing to take off. What works is another full coat over the whole area, mixed properly, applied wet edge to wet edge in one session. Pigmented coatings build opacity with thickness, so a sound extra coat over a streaked one usually hides it, provided the streaked coat is properly cured, abraded and within its recoat window. If it is outside that window you have an intercoat adhesion problem as well as a color problem, and that ordering matters: solve the bond question first, because a coat that hides the streaking and then delaminates has not fixed anything.

The first coat looked fine and the second one streaked. Why?

Because the second coat is doing a different job, and it is much less forgiving. A first coat over gray concrete is being judged against a substrate that hides a lot. Shade variation reads as normal texture in the slab, and nobody looks twice. The second coat is being judged against a uniform colored background, so any variation in it has nothing to hide behind. The same mixing discipline that passed unnoticed on coat one is suddenly visible. There is also a thickness effect. The second coat is usually applied thinner, and a thin pigmented film hides less of what is under it, so any variation in the first coat starts telegraphing through at the same time. The practical conclusion is that the mixing standard for a topcoat is higher than for a base coat, not lower, which is the opposite of how most crews treat the last coat of the day.

Does temperature or thinning cause this?

Both can, and both are worth ruling out because they change what you do next. Cold material is thicker and harder to disperse. The same paddle, drill and mixing time that produce a uniform batch at 70F will leave pigment in clumps at 50F, so a crew that mixes to a habit rather than to a result gets streaking on cold mornings and not on warm afternoons. Thinning is the more serious one. Adding solvent to a pigmented coating reduces the pigment per unit of film, so a thinned batch is genuinely a paler batch, and thinning by eye on some batches and not others guarantees a shade difference. It also reduces the film build, which compounds the hiding problem in the answer above. If a floor streaks only in certain conditions or only on certain days, the cause is usually one of these rather than the product.

Streaking in Solid Color: Why the Shade Is Uneven

A streaked floor is usually a mixing failure wearing an application costume. The bands look like they came from the roller, so that is where people go looking, but the pigment was already unevenly distributed before any of it reached the floor. Knowing which of the two you have decides whether the fix is a technique change or a recoat.

Quick answer

Scratch through a light band and a dark band. If the shade difference runs all the way down to the substrate, the pigment was unevenly distributed before the material reached the floor and the fault is in the bucket. If the film is the same shade through its depth and only the sheen differs, you are looking at roller and lap marks instead, which is a technique problem with a different fix.

Shade or sheen: the test that splits them

The difference runs through the film

Scratch to the substrate in a pale patch and a dark one. Both are the same thickness, different color
Cause: pigment distribution. It happened in the bucket, not on the floor
Follows the pour. Changes where one kit ended and the next began, and swirls rather than runs straight
This page. The fix is mixing discipline and usually another full coat

The difference is only how light comes back

Same shade at any depth. What varies is gloss, and it moves as you change viewing angle
Cause: application. Wet edge lost, uneven film, or a pass worked after it started to set
Follows the applicator. Runs parallel, at roller width, and lines up with pass starts and stops
Go to roller and lap marks

The pale areas are also soft

Press a fingernail into the lightest and darkest areas and compare
Cause: the mix was short of more than pigment. Unreacted material is soft material
This is a ratio and mixing failure that happens to show as color
Go to not curing or still tacky

Why under-mixed material looks mixed

Pigment is heavy, and it is carried in a paste that needs shear to break up rather than motion to move around. Stirring disperses the swirls long before it disperses the pigment, so the batch reaches a point where it looks uniform in the bucket and is not. That is the whole trap: the visual cue people stop on arrives several minutes before the job is done.

This is why mixing should be timed rather than judged. Most data sheets specify a mixing time and a speed, and for a pigmented system it is measured in minutes. Reading that number off the sheet is the single cheapest thing on this page, and how to read a data sheet covers where it sits and what else is next to it.

The four ways the pigment ends up uneven

Under-mixing, which is first by a wide margin. A paddle on a drill produces shear; a stick produces motion. Fold material from the bottom back through the middle, and run the clock rather than watching for the swirls to go
Settled pigment in storage. A kit that has stood for months carries a denser, more heavily pigmented layer at the bottom. Pour off the top and you have a pale batch now and a dark one later, from the same kit
Unscraped sides and bottom. Material clinging to the wall is both off-ratio and off-color, and it usually reaches the floor in the last few square feet of the batch. That is why the end of a pour streaks more often than the middle
Batch-to-batch inconsistency. Two kits mixed for different lengths of time, at different temperatures, or by different people are two different colors, even off the same production run

Notice that three of the four are about the last part of the batch rather than the first. A crew that mixes carefully and then scrapes the bucket out at the end has undone the mixing in the same movement.

Why the second coat is the one that shows it

A first coat is judged against gray concrete, which hides a great deal. Shade variation reads as texture in the slab and nobody looks twice. The second coat is judged against a uniform colored background, so the same variation has nothing to hide behind.

There is a thickness effect stacked on top of that. The last coat of the day is usually the thinnest, and a thin pigmented film hides less of whatever is under it, so variation in the first coat starts telegraphing through at exactly the moment the second coat stops covering for it. The film thickness calculator will tell you the spread rate a target build actually demands, which is worth checking before stretching a batch to finish an area.

The practical conclusion runs against habit: the mixing standard for a topcoat is higher than for a base coat, not lower.

Cold mornings and thinned buckets

Two conditions produce streaking on some days and not others, and both are worth ruling out before blaming the product.

Temperature. Cold material is thicker and harder to disperse, so the paddle, speed and time that give a uniform batch at 70F leave pigment in clumps at 50F. A crew mixing to a habit rather than to a result streaks in the morning and not in the afternoon
Thinning. Adding solvent reduces the pigment per unit of film, so a thinned batch is genuinely a paler batch. Thinning by eye on some batches and not others guarantees a shade difference, and it reduces film build at the same time, which compounds the hiding problem above

What actually fixes a streaked floor

The intuitive repairs all make it more obvious, so it is worth knowing why before reaching for them.

Spot-coating puts a third shade on the floor. Even a correctly mixed batch of the same product will not match a badly mixed one, so the patch reads as a patch
Buffing changes sheen, not color. If the difference runs through the film there is nothing to take off
Another full coat over the whole area is the repair that works, mixed properly and laid wet edge to wet edge in one session. Pigmented coatings build opacity with thickness, so a sound coat over a streaked one usually hides it

One ordering matters. That extra coat only holds if the streaked coat is properly cured, abraded and still inside its recoat window; outside it, you have an adhesion problem as well as a color problem. Settle that first with intercoat delamination, because a coat that hides the streaking and then lets go has fixed nothing.

Is it cosmetic?

Usually, with one exception worth checking before signing it off. If the streaking is purely pigment distribution, the resin and hardener were still correctly proportioned, the film cures and performs normally, and on a plant floor it may simply be acceptable.

The exception is when uneven color is a symptom of incomplete mixing rather than uneven pigment. Material that was not mixed long enough is also not fully reacted, and that shows up as soft or tacky areas rather than as color alone. So press a fingernail into the palest and darkest patches. If both are hard, it is an appearance defect. If either prints or stays tacky, it is a cure failure and the color is the least of it.

Still not sure which of these you are looking at? Start from the symptom and work back with the defect router.