One option: if you remove them from robots.txt, then Googlebot will crawl them, see that they are redirects, and not index them.

header('Location: http://www.example.com/?id=154077'); 
header('X-Robots-Tag: noindex, nofollow', true);

thank you for your help deadsea and levo. I prefer the levo's solution.
Once I add header('X-Robots-Tag: noindex, nofollow', true); will google recognize itself he has to remove it from the index or do I have to go to webmaster tool to remove it manually?

You just have to remove them from robots.txt. No manual action needed.

If they can't crawl it, they can't see those meta headers.

remove them from robots.txt

The issues with X-Robots noindex are exactly parallel to the issues with the meta robots noindex currently being discussed in detail in this thread...

Pages are indexed even after blocking in robots.txt
http://www.webmasterworld.com/google/4490125.htm [webmasterworld.com]

Yes, you must remove the pages from robots.txt. If you do that and use X-Robots noindex, you do not need to use the WMT removal tool.

Is the order important?
I mean:
Can I write
header('X-Robots-Tag: noindex, nofollow', true);
header('Location: http://www.example.com/?id=154077');
too?

the header order is not significant.

I have 48 php files for affiliation links. So I would include a file.php at the beginning of each redirect that contains header('X-Robots-Tag: noindex, nofollow', true); so next time I have to modify something then I only need to update only one file.

nofollow.php contains:
header('X-Robots-Tag: noindex, nofollow', true);

But when I try to modifiy redirect.php in this way:
<?php
include($_SERVER['DOCUMENT_ROOT'] . '/path/nofollow.php');
header('Location: http://www.example.com/?id=154077'
?>

It says:
header('X-Robots-Tag: noindex, nofollow', true);
Warning: Cannot modify header information - headers already sent by (output started at path/nofollow.php:2) in /home/path/public_html/redirect.php on line 4

that means I cannot use include to add header()?

Make sure nofollow.php file doesn't have any space/newline/other characters outside of <?PHP ?> tags.

Thank you levo. Now it works.
Is there a way to check if it is really nofollow?
On a static page I can do view source and check for
<META NAME="ROBOTS" CONTENT="NOINDEX, NOFOLLOW">, but here there is no static page.

Or do I have to wait for google's train?

You can use 'fetch as Google' in Google Webmaster Tools.

HTTP/1.1 302 Found
Content-Encoding: gzip
Vary: Accept-Encoding
Date: Sun, 09 Sep 2012 20:23:13 GMT
Server: LiteSpeed
Connection: close
X-Powered-By: PHP/5.2.17
X-Robots-Tag: noindex, nofollow
Location: [...........................]
Content-Type: text/html
Content-Length: 21

correct, right?

if you provide a 302 status code to googlebot then it will index the redirected content at the original url.
you need to provide a 301 response code and by providing a Location: header without specifying the status code it defaults to a 302.
therefore, your response must start with the 301 status code.

To do a 301 redirect change the location line to:

header('Location: http://www.example.com/?id=154077',TRUE,301);

Thank you for your help. This is the new one:

HTTP/1.1 301 Moved Permanently
Content-Encoding: gzip
Vary: Accept-Encoding
Date: Mon, 10 Sep 2012 07:52:34 GMT
Server: LiteSpeed
Connection: close
X-Powered-By: PHP/5.2.17
X-Robots-Tag: noindex, nofollow
Location: http://www.example.com/?id=154077
Content-Type: text/html
Content-Length: 21

Is that ok?

that looks good but it depends and what you ultimately want to happen.

i've never actually tried sending a X-Robots-Tag: noindex header with a 301 response, but i would assume that it will remove the requested url from the index and then not do a follow up request for the url in the Location: header.

if instead what you want to do is replace the requested url (which had been previously indexed) with the url and content from the Location: header then you should remove the X-Robots-Tag header and let the 301 show the search engine where to go.

My target is remove the redirect.php files from the index.
When I search for site: for my website it returns 48 php redirect files that I use as affiliation links to earn money. They should not appear in the index, because they are redirect to money links.

I do not why google shows such redirect.php files starting from a few days ago (they have been unknown for years) because they was on robots.txt and when I link I add rel="nofollow" ... to the <a> tag

Can someone tell me if I made the right move?
I have to edit 48 files and it will be a waste of time, so if you could confirm it would be a great help.

in that case i would go with what you have here (301 & X-Robots-Tag)

My php pages are disappeared after only 12 days from when I applied your advices.

It can take several months. Twelve days is good.

my rule of thumb is <2 weeks = fast & >2 months = slow but a lot depends on the number of pages affected vs indexed and the crawl rate.
a natural decay over time is not unusual.